Books, Scribes and Sequences in Medieval Norway Vol. 1 - Aaslaug Ommundsen

Åslaug Ommundsen

Volume 1: Parts I & II: Method and analysis

Dissertation for the degree philosophiae doctor (PhD) at the University of Bergen

2007

2

Acknowledgements First of all I want to thank the Centre for Medieval Studies at the University of Bergen (CMS) for granting me a three year position as a doctoral fellow. The key members, the administrative staff and the associates of the CMS have provided the best possible environment for a PhD-project. This thesis would not have been a reality if not for the invaluable support and patient guidence from my supervisor Prof. Lars Boje Mortensen, who has encouraged me to pursue the study of Latin philology since our first conversation during my first year at the University in 1992. At the CMS I also want to thank Dr. Jonas Wellendorf for reading large parts of my thesis and giving useful feedback and advice, as well as references to Old Norse primary and secondary literature. Dr. Aidan Conti has also contributed with advice on literature and palaeography. Prof. Else Mundal and Dr. Eldar Heide have contributed not only with insightful conversation, but also with housing and food respectively. Many thanks go to my second supervisor, my German “Doktorvater” Prof. Andreas Haug in Erlangen, who was the first to make me aware of the work of Dr. Lilli Gjerløw many years ago. I am very grateful for his support throughout this project, and particularly in the last, hectic phases of the dissertation. Dr. Wolfgang Hirschmann, the head of the Bruno Stäblein Archiv, also provided much needed help during my stay in Erlangen in 2004, for which I am grateful. During this project I have relied on the goodwill of the National Archives (Riksarkivet) in Oslo. Gunnar I. Pettersen has with admirable patience accepted my endless lists of requested fragments, not to mention the orders of photographs. I also want to thank Dr. Espen Karlsen for sharing the results from his work in the National Archives so generously with me. Yngve Nedrebø at Bergen State Archives and Ivar Oftedal in Stavanger State Archives both make it easy for people like me to get access to the smaller fragment collections. Bente Kvame at the library in Stavanger Museum has also been most helpful. In Bergen we benefit from a very professional and helpful staff at the University Library. I particularly want to thank Dr. Rune Kyrkjebø for involving me in the idea to present manuscript fragments on line, and for his 3

encouragement throughout my project. I also want to thank Tone Modalsli at the National Library in Oslo for her commitment for the medieval manuscripts and fragments, and for her always kind cooperation and interest. Monica Aase at Gunnerus’ library at NTNU received me in a very hospitable manner during my visit there. In Copenhagen Erik Petersen at the Royal Library is always encouraging and helpful, and an ally in matters concerning manuscripts. Furthermore I want to thank the members of the two international networks, first the international network of palaeographers in the Oslo workshops. Dr. Teresa Webber and Michael Gullick have offered help so generously that it has not been possible to credit them in every instance. I thank Michael Gullick for everything he taught me during his month at the CMS in the autumn 2006, and for his help in improving both the content and the language of my thesis. In the network of musicology and sequence research both Dr. Lori Kruckenberg and Dr. Gisela Attinger have been most encouraging and easy to turn to for help and advice, when a philologist has been lost in the mysteries of musicology. I want to thank Prof. Odd Einar Haugen for so clearly demonstrating the advantages of collaboration in the fields of Latin and Old Norse philology, and Alf Tore Hommedal (NIKU) for his kind help with questions regarding Archaeology. Dr. Kristin Sampson and Dr. Margareth Buer Søiland have showed me that a finished dissertation is worth the struggle, and have offered invaluable support. For making me aware of sequences as a living form of music, and for all the encouragement during this project I am grateful to all the singers in Schola Solensis and the conductor Halvor J. Østtveit. Last but not least I want to thank my family, particularly my sister Elisabet for patiently feeding fish, watering plants and taking general care of the Stavanger household for the last three and a half years. None of the people mentioned above can be blamed for any shortcomings in this thesis.

4

Contents 8

List of maps, tables and figures

10

Introduction

The arrival of writing Sequence studies in Norway A method for fragment studies Latin in Norwegian palaeography A genre study Centre-periphery Outline of the thesis

10 11 13 14 16 18 19

Part I: To study manuscript fragments 1. Research history 1.1. Norwegian fragment studies

24

1.2. The palaeographical traditions

31

1.3. The sequence: rise, fall and resurrection

41

1.4. The archbishopric of Nidaros

55

1.1.1. The first focus 1.1.2. Liturgy as law 1.1.3. New initiative 1.1.4. Fiefs and provenance

24 25 27 28

1.2.1. The purposes of palaeography 1.2.2. Derolez and Scandinavia 1.2.3. Old Norse and Latin palaeography in Norway 1.2.4. Codicology without the codex 1.3.1. The sequence – history, form and contents 1.3.2. International sequence research 1.3.3. Scandianvian sequence research 1.3.4. A new book on the Nidaros sequences 1.4.1. Nidaros 1.4.2. Bishop’s sees, monasteries and scriptoria 1.4.3. The date of the Nidaros ordinal

31 32 34 38

41 45 47 49 55 57 66

2. The remains of a book culture 2.1. Remains and losses - an estimate

68

2.2. Are the liturgical manuscripts representative?

82

2.1.1 What is left? 2.1.2 What was once here? 2.1.3 How much is lost, and what are the consequences? 2.2.1. Liturgical manuscripts as books 2.2.2. Fragments as manuscripts 2.2.3. The scribes

69 72 79 82 83 86

3. Approach and method 88

3.1. Selection

3.1.1. Genre study 3.1.2. Manuscripts, scribes or scriptoria 3.1.3. Manuscripts within a timeframe

88 89 91

5

3.2. Searching for European centres in Norwegian fragments 3.2.1. Dividing the local from the imported 3.2.2. Localising regional European traits

92

93 95

3.3. Grouping manuscript fragments

105

Summary and conclusion to Part I

108

Part II: Analysis 4. The twelfth century 4.1. A sequence for mass and magic: Seqv 29 (Br-mi 2), s. xii¹

114

4.2. A grand-scale missal for a grand-scale church: Seqv 31 (Mi 80), s. xii¹

122

4.3. Two sequentiaries of different origin: Seqv 40 and 53 add, s. xii med.

129

4.4. Norwegian sequentiary - or English missal? Seqv 11 (= Mi 133?), s. xii²

134

4.5. German sequences in Anglo-French script: Seqv 30, s. xii²

142

4.6. A priest’s manual in German script: Seqv 59b add (ManNor F), s. xii²

147

Conclusion for the twelfth century

151

4.1.1 A twelfth century breviary-missal 4.1.2. From France to Norway 4.1.3. Alma chorus in Norway

4.2.1. English scribe, German neumes and punctuation 4.2.2. A product of Lund? 4.2.3. A German sequence 4.2.4. Further research 4.3.1. An English sequentiary 4.3.2. A Norwegian sequentiary 4.3.3. A mixture of English and German sequences 4.4.1. A scribe of unknown origin 4.4.2. One book or two? 4.4.3. Circumstantial evidence 4.4.4. Norway or England?

4.5.1. German sequences for Easter and the Ascension 4.5.2. “Anglo-French” script 4.5.3. A local product of mixed influence 4.6.1. Before 1200 4.6.2. Germany, Denmark or Norway? 4.6.3. A Notker-sequence

114 117 118 122 124 126 127

129 131 133 134 135 137 140 142 144 145

147 149 149

5. The thirteenth century 5.1. The missal and the Old Norwegian Homily Book: Seqv 39b add (Mi 38), s. xiii¹

157

5.2. On the track of an unknown local scribe: Seqv 51 add, s. xiii¹

172

5.3. A missal of the highest standard: Seqv 33b add (Mi 107), s. xiii¹

181

5.1.1. An unusual sequence 5.1.2. Two liturgical manuscripts – one scribe 5.1.3. One scribe or many? The problem of the Old Norwegian Homily book 5.1.4. A single scribe or a scriptorium? 5.1.5. Suggesting an origin 5.2.1. Scribal characteristics 5.2.2. A third manuscript

5.3.1. A Norwegian scribe writing in English style 5.3.2. Possible models 5.3.3. Sequences not secundum ordinem

6

157 158 160 164 168 172 177

181 183 186

5.4. A manuscript with unknown sequences: Seqv 25, s. xiii¹

188

5.5. Three local sequentiaries – and one roll: Seqv 1 et alia, s. xiii¹

193

5.6. Manuscripts secundum ordinem: Seqv 18 and 38 add, s. xiii med

201

5.7. The St. Olav scribe: Seqv 45b add and Seqv 64 add, s. xiii²

213

Conclusion for the thirteenth century

219

5.4.1. A sequence for St. Andrew? 5.4.2. Quare fremuerunt for the Holy Blood? 5.5.1. Seqv 1 5.5.2. Seqv 2 5.5.3. Seqv 15 5.5.4. A roll

5.6.1. A missal and a breviary-missal 5.6.2. Grates honos hierarchia 5.6.3. Liturgy of summer and early autumn 5.6.4. The authority of the Nidaros ordinal 5.7.1. Six books from one hand 5.7.2. Not beautiful, but confident

188 189

193 195 197 198

201 205 207 211 213 215

Conclusion

A book-culture of diversity Sequences in context Ad fragmenta!

222 227 229

Indices

233

Bibliography

258

Manuscripts Gjerløw’s alternative sigla for manuscripts with sequences Sequences in Norwegian and Icelandic fragments and manuscripts Contents of Norwegian manuscripts with sequences Sequences through the church year

7

234 239 240 247 253

List of maps, tables, figures and diagrams

Maps: Map 1: Map of the Norwegian fiefs c. 1640. Map 2: Main land bishoprics and religious houses in medieval Norway. Map 3: The Nidaros archbishopric and a selection of European centres.

30 63 152

Tables: Table 1: Periodisations and terminology in Latin and Old Norse palaeography Table 2: Religious houses in medieval Norway Table 3: Collections holding selected Latin and Old Norwegian books Table 4: Estimated numbers of books in Norway in the Middle Ages and today Table 5: Twelfth century manuscripts with sequences Table 6: Sequences of Easter-week and the Ascension Table 7: Thirteenth century manuscripts with sequences Table 8: The contents of Seqv 1 (Lat. fragm. 418) Table 9: Seqv 1 and Seqv 2 and selected English and French sequentiaries Table 10: Selected manuscripts with sequences from June to August

35 64 72 79 113 143 155 194 196 208

Figures: Fig. 1: Seqv 29 (Br-Mi 2), NRA Lat. fragm. 251-2r. 115 Fig. 2: Seqv 29 (Br-mi 2), NRA Lat. fragm. 251-2v: Alma chorus domini. 118 Fig. 3: Parchment amulet. Alma chorus. Aust-agder kulturhistoriske senter. 120 Fig. 4: Seqv 31 (Mi 80), Trondheim, Gunnerus’ Library, fragm. 9. 123 Fig. 5: Lund University library, MS 5, f. 150v. 125 Fig. 6: Seqv 40 add, NRA, Lat. fragm. 776. An English sequentiary in Norway. 130 Fig. 7: Seqv 53 add, NRA. Lat. fragm. Sandaakers kat. LR pk 142-1. 132 Fig. 8: Seqv 11, NRA, Lat. fragm. 497: Fulgens preclara. 136 Fig. 9: Mi 133, NRA Lat. fragm. 354-1 (new number: Fr.703). 136 Fig. 10: Mi 133 (NRA Lat. fragm. 354-1, new number: Fr.703). The size is reduced. 139 Fig. 11: To the left: A segment from Seqv 11 (Lat. fragm. 497). To the right: A segment from Seqv 16 (Lat. fragm. 471). 140 Fig. 12: Seqv 30, NRA Lat. fragm. 236: Summi triumphum. 145 Fig. 13: Seqv 59b add, Copenhagen, Royal Library NKS 32 8°, f. 34r. 148 Fig. 14: Oslo, NRA, Lat. fragm. 764 (Mi 38/Seqv 39b add): Gaudia mundo ventura (inc.). 158 Fig. 15: Above: Oslo, NRA, unnumbered leaf, Salten 1635 (Ant 7): St. Olav’s office In regali fastigio. Below: Oslo, NRA, Lat. fragm. 764 (Mi 38). 159 Fig. 16: “Hand 1” of the Homily book. Above f. 2r, below f. 16r 162 Fig. 17: “Hand 2” of the Homily book, f. 36r. 162 Fig. 18: “Hand 3 and 4” of the Homily book, on f. 71v. 163 Fig. 19: Left (colour): Oslo, NRA, Lat. fragm. 764 (Mi 38/Seqv 39 add). Right (bl&wh): Selected initials from the antiphoner Ant 7. 165 Fig. 20: The M to the left is from the inserted leaves in the Homily book. 166 Fig. 21: The M below to the right is from the main part of the Homily book. 166 Fig. 22: The small segments to the left: Seqv 51 add (colour) and Lec-Mi 4 (bl-wh). 174

8

Fig. 23: To the right: Segment of Seqv 51 add (Oslo, NRA, Lat. fragm. Sandaakers kat. LR, pk. 157). 174 Fig. 24: A copy of Lec-Mi 4 (Oslo, NRA, Lat. fragm. 452). The reproduction is not in 1:1. 175 Fig. 25: Gr 27 (NRA, Lat. fragm. 433). 178 Fig. 26: Seqv 33b add (NRA Lat. fragm. 489, new number: Fr.628). 182 Fig. 27: Seqv 33b (NRA Lat. fragm. 489-11, new number: Fr.614). 184 Fig. 28: Stavanger, statsarkivet, fragm. no 1-3 (Mi 35). 185 Fig. 29: Oslo, NRA Lat. fragm. 103. Thirteenth century psalter. 186 Fig. 30: NRA Lat. fragm. 532-2. 189 Fig. 31: These lines from Seqv 2 (Oslo, NRA Lat. fragm. 336). 196 Fig. 32: Seqv 18/Mi 106 (Oslo, NRA Lat. fragm. 882-5, -6): Virgini marie laudes (inc.) 204 Fig. 33: Seqv 18/Mi 106 (Oslo, NRA Lat. fragm. 881-6, -7): Grates honos hierarchia. 207 Fig. 34: Seqv 38 add (Br-mi 3) (Oslo, NRA Lat. fragm. 668). 210 Fig. 35: Oslo, NRA Lat. fragm. 1031 (Br 1). 216 Fig. 36: Oslo, NRA Old Norse fragm. 47c, King Magnus Lagabøte’s Lawbook. 217 Fig. 37: Oslo, NRA Lat. fragm. 1031 (Br 1). 217

Diagrams: Diagram 1: The chronological distribution of manuscript with sequences Diagram 2: The origin of twelfth century manuscripts with sequences Diagram 3: The origin of fifteenth century manuscripts with sequences Diagram 4: The genre distribution of the manuscripts with sequences

9

222 226 226 227

Introduction The basic source material of this study is a selected group of manuscript fragments of Norwegian provenance containing a particular type of medieval chant, generally referred to as “sequences”.1 The emphasis here is not primarily on the sequences, but on the fragments transmitting them and the physical evidence of books which once existed in Norway in complete form. This thesis explores how the fragments with sequences can be used to study a growing manuscript culture in the periphery and its innate European influences. The study of manuscript fragments is of immense importance in Norway to increase our knowledge of medieval book and scribal culture, as so little material is transmitted in the form of complete codices. In spite of this the fragments have attracted little attention, especially in the field of Latin philology. In this study all Norwegian fragments with sequences are described in an illustrated catalogue. In addition, selected items are analysed to increase the knowledge of book and scribal culture in Norway and the influences upon it from other European regions in the twelfth and thirteenth century. The study also adresses the general challenge of fragment studies, and seeks to provide answers to the following questions: How can fragments from Latin manuscripts, particularly liturgical, best be approached in a study of medieval book culture? How can studies of such fragments shed light on the cultural transfer between European centres and the northern periphery in the Middle Ages?

The arrival of writing In the northern regions of Europe the history of the book more or less began from scratch with the introduction of Christianity around the turn of the first millennium.2 The need for liturgical books as the new religion took hold on a large scale was an important reason to establish local production centres and there is some foundation 1

The genre is also known as “prosa”, and the melodies referred to as “sequentiae”. This unique feature of the northern (and partly eastern) regions of Europe was for instance pointed out by Michael Gullick at a workshop in Bergen October 2005 (cf. Ommundsen 2006a, 42-43).

2

10

Introduction for claiming that liturgical books represent the very beginning of a Nordic book and 3

scribal culture in the Middle Ages. So far the potential of early liturgical books to reveal information about the arrival of writing, Nordic book culture and its dependency on European centres has not been fully explored. The fundamental notion behind this study is that liturgical manuscripts, mainly preserved in the form of fragments, can supply information that goes beyond the scope of liturgical or musicological studies and into the history of the book in the North.4

Sequence studies in Norway At the core of this study is a selected group of fragments from liturgical manuscripts once used in Norway, all containing a particular type of song used during the medieval Mass: the sequence. Sequences were songs of celebration composed for particular feastdays, and although they were not part of the original Roman liturgy, they became very popular and were produced in large numbers all over Europe. An international project involving a group of musicologists has recently focused on the sequence repertories of Scandinavia, Nidaros in particular, on the initiative of Andreas Haug (Erlangen/Trondheim).5 This study has benefited much from collaboration with this group. The results of the project have now been published in a book called The Sequences of Nidaros: A Nordic repertory & its European context (Kruckenberg and Haug 2006), which also included two of my case studies (Ommundsen 2006b).

3

The importance of the mass books in the early stages of the Nordic scribal culture is based on the results from the first international workshop at the National Archives, Oslo, held in August 2003 (Karlsen 2003). 4 That studies of medieval book history in Norway to a large degree depends on the fragmented remains of manuscripts in the Norwegian National Archives (NRA) is emphasised by Espen Karlsen in the publication of the results from the first international NRA workshop in 2003 (Karlsen 2003, 58). As Karlsen points out, Sverre Bagge also recognised the importance of the NRA fragment collection in Norsk idéhistorie, vol. 1: Da boken kom til Norge (Bagge 2001, 80). In addition the importance of the fragments as sources of book history is emphasised in several Swedish publications in connection to the MPO-project (Medeltida PergamentOmslag) at the Swedish National Archives (Abukhanfusa, Brunius, and Benneth 1993; Abukhanfusa 2004; Brunius 2005). 5 The project called “The Sequence Repertory of Nidaros within its European Context”, funded by the Norwegian Research Council, has been organised by Andreas Haug and Lori Kruckenberg. Meetings were held in Trondheim 2003, Yale 2004 and München 2005.

11

Anyone in Norway working with sequences and the fragments transmitting them owes much to Erik Eggen (1877-1957) and Lilli Gjerløw (1910-1998), who each identified a large number of fragments with sequences in the Norwegian National Archives in Oslo (NRA). Eggen’s work on the sequences in Norway and Iceland was published after his death as The sequences of the archbishopric of Nidarós, edited by Jón Helgason, in Bibliotheca Arnamagnæana (1968). The same year Gjerløw published the book with which all studies in the liturgy of Nidaros should begin, namely Ordo Nidrosiensis ecclesiae (ON), an edition of the official Nidaros ordinal issued in the beginning of the thirteenth century. The edition came with an appendix devoted to sequences (Gjerløw 1968, 431-439). While Eggen discovered fragments from thirty manuscripts with sequences, Gjerløw later found fragments with sequences from more than thirty additional manuscripts, and listed a total of sixty two in her unpublished catalogue of liturgical manuscripts, often with tentative dates.6 Until now these additional sources have not been available outside the NRA, and therefore not easily accessible to scholars. In practice, this has limited the research on sequences to the group of fragments identified by Eggen. This will hopefully change in the future, as now all identified Norwegian manuscripts with sequences are available in an illustrated catalogue for the benefit of manuscript studies and musicological studies alike. The catalogue, which is organised in correspondence with Gjerløw’s catalogue numbers, forms an independent section of this thesis.7 Fragments from two more manuscripts have been added, making the total number of entries in the catalogue sixty four. Both Eggen and Gjerløw sometimes included fragments from different manuscripts under the same catalogue number, and, in one instance, fragments from the same manuscript were given two different catalogue numbers. Although there are only sixty four

6

In her catalogue, Lilli Gjerløw also included some manuscripts which were not in Norway during the Middle Ages, and are therefore not immediately relevant to Norwegian book culture. These will be included in the catalogue, but excluded from the analysis. One fragment is lost, and one is privately owned. These have retained their numbers and positions in the catalogue, but are not described. 7 In the making of the catalogue I have consulted several catalogues and guidelines, like the Richtlinien Handschriftenkatalogisierung (Deutsche Forschungsgemeinschaft 1992), Medieval manuscripts in British Libraries (Ker 1969) and the NRA database. The benefits of illustrations are demonstrated not only by Eggen and Gjerløw, but also in the series of Illustrated inventories of Medieval Manuscripts (Gumbert 1991)

12

Introduction catalogue numbers, some are sectioned into a and b, making the total number of manuscripts in the catalogue closer to seventy. The descriptions and estimates of date and origin in the catalogue will supplement Eggen’s work as his primary concern was the text and melody of the sequences and not the manuscripts the fragments were once part of nor their palaeographical and codicological characteristics. So, although sequences in Norway have been studied for a century, the manuscripts transmitting them have, with a few exceptions, not been thoroughly described or analysed, and this current study seeks to remedy that.

A method for fragment studies The purpose of this study goes beyond simply presenting unpublished sequence material and describing it. It seeks to explore how liturgical manuscript fragments can be approached in studies of book culture, and be used to trace influences from the European centres which played a role in a place like Norway on the northern borders of Europe, both through imported books and books which were locally produced. The present author is a Latin philologist, and not a historian or a musicologist. Still, this study is no more philological than it is musicological, but is rather an attempt to take palaeography and codicology past the level of “auxiliary sciences” to something with independent value. Both sciences have the potential to go beyond simply being tools to make a manuscript catalogue or to achieve a correct transcription or better understanding of a transmitted text. In that sense this study goes beyond ”New Philology” or ”Material Philology”;8 the text (or in this case, the sequence) is no longer the main purpose, but the book is, the scribe who wrote it, and the culture that the book and scribe were both part of. The contents are used to gain a better understanding of a given book, its origin and its role in a growing book culture on the northern edges of Europe, as the books themselves are testimonies to human cultural activity and cultural exchange in the Middle Ages.

8

cf. the discussions of New Philology in Speculum vol. 65, No. 1 1990, recognising both the physicality and diversity of the medieval manuscript culture (Nichols 1990) and the importance of broad contextual information to elucidate a text (Wenzel 1990).

13

How this can be done has hitherto not been described in writing, and in the first section of this study I will discuss some methodical considerations and guidelines to distinguish significant scribal features, and identify the letters and signs which may be used to identify imports or influences from different European regions. By selecting manuscripts with sequences dating from the twelfth and thirteenth century for a closer analysis, using the guidelines drawn up in the first part of the study, I hope to show how it is possible to recognise which regions of Europe these early books were closely connected to and dependent upon.

Latin in Norwegian palaeography The method outlined in the first section is based on palaeography, which in Norway has traditionally been closely linked to Old Norse philology. In his fundamental description of Norwegian and Icelandic palaeography, Palæografi B. Norge og Island, the Norwegian linguist and palaeographer Didrik Arup Seip (1884-1963) mentioned the possibility that some of the liturgical manuscripts with musical notation could have been produced in Norway, but expressed uncertainty on this point and ventured no further (Seip 1954, 68). Later descriptions of the development of Latin script in Norway have not included Latin texts, due to the difficulty of identifying scribes of Norwegian origin without the aid of language (Haugen 2002, 825; 2004, 177). Norwegian scribes writing books in Latin have for the main part been hidden behind the anonymity of the Latin language, unless the contents of a book revealed its northern origin. It is certainly ironic that manuscripts in Latin have been considered unfit for the study of Latin script. One scholar during the last century ventured to identify Norwegian scribes in fragments from liturgical manuscripts, Lilli Gjerløw,9 but unfortunately her work has had little impact outside the field of musicological and liturgical studies.

9

See in particular Ordo Nidrosiensis Ecclesiae (Gjerløw 1968), Antiphonarium Nidrosiensis Ecclesiae (Gjerløw 1979), “Missaler brukt i Bjørgvin bispedømme...” (Gjerløw 1970), “Missaler brukt i Oslo bispedømme...” (Gjerløw 1974).

14

Introduction Although the challenges involved in recognising Norwegian scribes writing in Latin are considerable, developments in recent years have shown that it is possible to distinguish locally produced material from the imported with some degree of confidence. In connection with a project to register the liturgical fragments in the NRA in a database, three workshops were organized involving an international network of palaeographers.10 With the help provided in Gjerløw’s work and through these international workshops, it has been possible to begin a closer study of the remains of Latin books and to try and find answers to the following questions: • Which books were imported, and from which part of Europe – and which books were locally produced? • From where were the Norwegian scribes influenced? Who were their teachers and what models or exemplars did they have and use? The indications from the international workshops are that although English influence has been strong, the earliest “local” fragments, possibly testifying to a scriptorium in the Oslo or Vika area in the late eleventh century, show a mixture of German and English influence (Karlsen 2003, 66-70). Although it has long been assumed that early script was also influenced from Germany and/or Denmark, it has been difficult to trace evidence of this in vernacular script or palaeography (Seip 1954, 3; Haugen 2002, 826). The traditional approach to script development in Norway has been very much focused on English influence, which is natural due to a close contact with England during the process of Christianisation, and the insular letters in the vernacular alphabet. At some point, however, one may say that the Old Norse alphabet was relatively established (although the use of letters varies a little) and that the insular 10

The workshops at the NRA are due to the initiative of Espen Karlsen (Trondheim/NRA), Gunnar I. Pettersen (NRA) and Andreas Haug, and were arranged in August 2003, March/April 2005 (with the cooperation of CMS, Bergen and the centre for medieval studies at NTNU, Trondheim) and August 2006. The international network of scholars includes Dr. Teresa Webber (Cambridge), Prof. David Ganz (London), Michael Gullick (Walkern), Dr. Susan Rankin (Cambridge), Dr. Christian Heitzmann (Wolfenbüttel) and Dr. Gunilla Björkvall (Stockholm). A fourth workshop was organized by the CMS in Bergen in October 2005 to discuss the status and general problems of fragment studies in the Nordic countries.

15

letters represent an older stage of influence in script development. Although several letterforms as described by Seip in his Palæography are used in English script, they are also generally used in other regions of Europe and can, perhaps with the exception of the tall a (Seip 1954, 14), not in themselves be used as arguments for English influence after a certain point. If we assume that Norwegian scribal culture was constantly facing new models and new influences from the rest of Europe after the establishment of a scribal culture, the influence on script over time and in different regions and scriptoria must be found in the execution and shapes of letters other than those borrowed from the insular alphabet. Significant letters can even be easier to recognise in a Latin text written by a Norwegian scribe than a vernacular one. The results of the current study indicate that the spectrum of influences on Norwegian scribes has been broader, and extended over a longer period than previously assumed. Through this study I wish to demonstrate that fragments from the Latin manuscripts have a natural and important place in Norwegian palaeography.

A genre study As will be discussed in the first section of this study, there are several ways to select material for manuscript or fragment studies, one of which is the selection of a book genre. There are several genres to choose from both among the non-liturgical and liturgical books. Books with liturgical contents may have the advantage of being produced in accordance with a recognisable rite of a specific diocese or institution. In studies of medieval Norwegian liturgy and music the focus has often been on the Office rather than the Mass. Larger Norwegian studies of considerable importance have in recent years focused on Offices or the Office chant books, like Antiphonarium Nidrosiensis ecclesiae (Gjerløw 1979), A Comparative Study of Chant Melodies from Fragments of the lost Nidaros Antiphoner (Attinger 1998), The Office of Saint Olav (Østrem 2001) and The Nidaros Office of the Holy Blood (Attinger and Haug 2004). Studies in the medieval Mass are often limited to sequences or a related phenomenon

16

Introduction called “tropes”, another type of embellishment of the original Roman liturgy introduced in the North. That tropes were used in medieval Norway is shown by the NRA fragment material, which contains, for instance, the Kyrie-trope Kyrie fons bonitatis in Lat. fragm. 418 (Seqv 1) and the trope Ab hac familia for the Offertory Recordare virgo mater in Lat. fragm. 1029 (Seqv 46 add). 11 Liturgical books, whether they were imported or locally produced, are tangible and specific results of impulses and influences from different European centres at different times, impulses which are traceable in form and contents alike. In this study the manuscript will both be seen as physical evidence of book production, and as a bearer of liturgical texts and music.12 The presence of one or more sequences in a fragment or manuscript has governed the selection of the studied corpus, as the manuscripts with sequences have particular advantages: the large variety in the European sequence repertories facilitates the identification of the regional origin or influence of the manuscript. A sequence or a combination of sequences may point in the direction of specific parts of Europe or particular liturgical uses. Since the sequence was an element in the Mass, the sequences are found in several kinds of books containing the sung elements in Mass celebration: the gradual (containing the chants for the Mass), the missal (containing all elements for the Mass, chants, readings and prayers), or the breviary-missal (containing all elements for both Mass and Office). In some instances, sequences were used instead of hymns in the celebration of the Office, and therefore in a few instances they are also found in antiphoners or breviaries.13 All these different book genres with sequences are represented here, and this study is therefore not a book genre study in its strictest sense. The selection of the corpus is, however, governed by one specific genre being present in the contents. 11

In her catalogue Lilli Gjerløw has registered only one “troparium”, namely Oslo, NRA Lat. fragm. 867, 1-3. It is likely that tropes have not been deliberately sought out in the fragment material of the NRA, and that more tropes exist in the unregistered fragment material. 12 Cf. the discussion of material codicology in Gumbert 2004. 13 For a brief and popular survey of the different liturgical books, see Helander in Abukhanfusa a.o. 1993, 106. For a broader instruction, see for instance Les livres de chant liturgique (Huglo 1988) or Medieval Manuscripts for Mass and Office: A guide to their organization and terminology (Hughes 1982).

17

With the exception of two codices and one parchment roll, all the manuscripts with sequences in this study are in a fragmentary state. In most cases only a couple of fragments have survived from each book, often not enough to reconstruct a single leaf, and fragments can only in rare cases be pieced together to create something resembling the original manuscript. Nevertheless, the fragments can still be seen as evidence and representatives of manuscripts, or more precisely, once-complete manuscripts.

14

Centre-periphery The archbishopric of Nidaros was established in 1152/53, with its centre in the town of Nidaros, current Trondheim. Nidaros shared its status with other northern and eastern European regions as part of a peripheral rim around the central and western parts of Europe. It quickly developed into an important centre in this periphery, receiving impulses and passing them on to smaller centres in its vicinity. Impulses also came to Nidaros via smaller Norwegian centres, and not only through the larger centres abroad, as indicated by the first chapter in the analysis. Nidaros archbishopric comprised Norway, Iceland, Greenland, the Faroe Islands, Shetland, the Orkneys, Sudor and Man. Still, this study is limited to mainland Norway. To include Iceland and deal with the changing relationship between the two countries and the rest of Europe during several centuries would make the study too complex. Iceland and Norway, even though united in the arch see of Nidaros, were two cultural entities during the Middle Ages, and the Icelandic material requires a study of its own.15 The other provinces outside of the Norwegian main land will also be excluded. Bohuslän in present day Sweden, however, will come under the scope of this investigation, since it was regular Norwegian mainland territory in the Middle Ages, and under the Nidaros arch see. The same is not the case for the regions Jemtland and Herjedalen, which, even though they were subject to the Norwegian king, belonged to the diocese of Uppsala. 14 15

I owe the term ”once-complete manuscripts” to Anja Inkeri Lehtinen (in Brunius 2005). Gisela Attinger is currently studying the Icelandic manuscripts with sequences (Attinger 2006).

18

Introduction

Results from this study show that it is possible through a limited number of twelfth and thirteenth century liturgical fragments to identify visible signs of the contact between medieval Norway and other European centres or regions. Different manuscripts with sequences may be argued to be physical testimony to contact with the archbishopric of Lund (Denmark), and with identified or unidentified centres in Germany, France and England.

Outline of the thesis The study is divided into three parts: Part I is concerned with research history, the surviving corpus of the medieval manuscripts, and methodical considerations. The chapters on research history focus on fragment studies, palaeographical traditions, international and local sequence research, and finally, past and current research on monasteries and scriptoria. The research history is followed by an investigation of the remaining evidence of the medieval manuscripts, losses of the medieval books, and the representativity of the remaining fragments, liturgical fragments in particular. This leads up to a discussion of approaches and methods with which one may study a medieval book culture through liturgical fragments. Part II provides the analysis of selected manuscripts from the twelfth and thirteenth century, presented more or less in chronological order. Part III is an illustrated catalogue with all the registered manuscripts with sequences in Norwegian collections from c. 1100-1500, which will also serve as a reference-work for the closer manuscript analysis in Part II.

19

20

Part I: To study manuscript fragments

Part I: To study manuscript fragments In most other countries “manuscript fragments” would simply be a subchapter to the major headlines devoted to manuscripts or medieval codices. In Norway, at least as far as Latin manuscripts are concerned, the situation is the other way round. To study manuscripts in practice means to study manuscript fragments, with the challenges this entails. Most of the manuscripts used in medieval Norway have vanished without a trace, mainly due to neglect and a general lack of interest in the centuries following the 16 Reformation in 1536/37. For the Latin manuscripts the situation is worse than for

those written in the vernacular. Only a handful of the medieval Latin manuscripts have survived in the form of codices. Instead we are mainly left with the 5-6000 single fragments from c. 1200 medieval manuscripts, most of which are liturgical, in the collection at the NRA in Oslo, and in some smaller public collections.17 Some Norwegian material is also kept in foreign collections, particularly in the Royal Library and the Arnamagnean Collection in Copenhagen. In the following chapters I will first adress the research history, since important work has been done in the field of fragment research, palaeography and codicology, sequence studies and the medieval context of books, which have benefited this study and serve as a foundation for its results. In the second section I will discuss how much remains and how much is lost of the original book corpus in medieval 16

The few cases of demonstrative book-burning in the aftermath of the Reformation probably had a smaller impact on the total corpus of liturgical books than previously assumed, cf. Ommundsen 2006b. Several books kept in book-collections and libraries in the larger towns have been lost in large town fires, particularly in Bergen 1623, Oslo 1624 and Copenhagen 1728 (Holtsmark 1956a). See also a list of the fires in monasteries in Holm-Olsen 1990, 149. 17 The estimated number of Latin manuscripts is based on the c. 1300 envelopes of Latin fragments and c. 100 of Old Norse fragments in the NRA, since at the time fragments in the bindings were detached from the accounts, fragments which looked like they came from the same manuscript were put in the same envelope. Since fragments from the same manuscript have been found in different envelopes, the number of manuscript is probably somewhat lower than the actual number of envelopes. The same estimation (1200 manuscripts) is also used by Odd Einar Haugen (Haugen 2002, 2004). There are some smaller public collections apart from the NRA also holding manuscript material from the Middle Ages, especially the Regional State Archives, particularly Stavanger and Oslo, larger libraries, like the National Library in Oslo and the University Libraries of Bergen and Trondheim and some local museums. These institutions hold a total of between two and three hundred single fragments. For a survey of these collections and their holdings, see Ommundsen 2006a, 59-61.

21

Norway. The purpose is to get a perspective on the remaining evidence and estimate how large (or small) a part of the original books is actually available to modern scholars. Finally, and most importantly, I will present practical approaches for studies of the existing manuscript material, and a general method for evaluating the origin or regional influence of manuscripts. This method will be applied to the selected corpus of manuscript fragments.

22

1. Research history

1. Research history This chapter on research history will be divided in four main sections, one for each of the four fields most relevant to this study: Fragment studies, palaeography, sequence studies and history/archaeology. Norwegian fragment studies have a history of more than 150 years with changing interests and priorities. The focus on Norway does not mean that the useful work done in other parts of Europe on fragment studies is disregarded. Quite the contrary, Norwegian fragment studies have been pushed forward by other efforts, particularly the Swedish work done during the last decade, and the growing awareness of the value of fragments in the rest of Europe.18 The second focus is on palaeography and codicology, which are the most vital tools in fragment studies, but which strangely enough in many cases have not been applied in the history of fragment research. Thirdly, as this study relates to earlier sequence research, the nature of the sequence and the modern research is presented in a separate chapter. Both palaeography and sequence studies have a national tradition linked to the international development of the field, and both will be considered here. Finally, one important goal for the future is to be able to relate as many manuscripts as possible to a specific historical context and connect them to particular centres. Although this in most cases may prove difficult, the first step is to take a closer look at the centres in Norway where books would be kept and/or copied. The last chapter will give an overview of the most important Norwegian bishops’ sees and religious houses, as well as the recent work on Norwegian scriptoria from an archaeological point of view.

18

For the now completed project to make a database of the fragments in the Swedish National Archives in Stockholm, referred to as the MPO-project, see Abukhanfusa, Brunius, Benneth 1993, Abukhanfusa 2004, Brunius 2005. Articles drawing attention to the value of fragments have appeared in the south of Europe as well as in the north. Merete Geert Andersen’s “Colligere fragmenta ne pereant” about fragments from Icelandic manuscripts in the Arnamagnæan Collection (Andersen 1979) shares its title since 2001 with the article “Colligere fragmenta ne pereant. Il recupero dei frammenti liturgici italiani,” stressing the need for fragments to supplement the relatively small corpus of transmitted northern Italian tenth to twelfth-century liturgical books (Baroffio 2001).

23

Part I: To study manuscript fragments

1.1. Norwegian fragment studies To select the corpus of sequence manuscripts as a topic for this study was possible because scholars at an early date took an interest in the fragment material and did the groundwork of identifying manuscripts with sequences. In the following pages I will give a brief sketch of the fragment collection in the NRA and the scholarly interest that has surrounded it until this point.

1.1.1. The first focus The parchment fragments at the NRA were first discovered by Henrik Wergeland in the early 1840’s as bindings of sixteenth and seventeenth century archival material (cf. for instance Munch 1847, 25).19 The different stages in detaching the fragments from the accounts and organising the fragment collection in the NRA in the following years are described by Pettersen (2003, 45-56).20 The scholarly focus on the fragment collection has been goverened by changing historical circumstances and priorities. When the first leaves were discovered in the NRA in the 1840’s, what exited the Norwegian historians like P. A. Munch (1810-63) were the Law codices and Saga texts written in Old Norse.21 Regarding the Latin fragments, Munch wrote: Blandt det saaledes Forefundne er vistnok det Meste Brudstykker af latinske Bøger af asketisk Indhold, og saaledes ei af nogen synderlig Interesse for os, uden forsaavidt de i Almindelighed ere meget smukt skrevne og synes at være ældgamle (Munch 1847, 26) 19

For the phenomenon of leaves from medieval manuscripts used as covers or bindings for accounts or tax ledgers of paper, see Abukanfusa, Brunius, Benneth 1993 or Abukhanfusa 2004 for the Swedish material, or Pettersen 2003 for the Norwegian material. 20 The same study (part I) along with Karlsen’s contribution (part II) contains valuable information also regarding other aspects of Norwegian fragment research and the history of the fragment collection (Pettersen and Karlsen 2003). 21 Fragments with Old Norse laws were published with facsimiles in Norges Gamle Love (the Old Laws of Norway) (Storm 1885), and some saga fragments were published in Otte brudstykker af den ældste saga om Olav den hellige (Eight fragments from the oldest saga of St. Olav) (Storm 1893). Fragments from an Old Norse translation of the Rule of St. Benedict and the Lanfranc was first edited by Ernst Walter in “Die Fragmente zweier Klosterregeln für Benediktinermönche in altnorwegischer Übersetzung” (Walter 1960), later by Lilli Gjerløw in Adoratio crucis (Gjerløw 1961) and in the supplements of the KLNM (XXI) under “Benediktinorden” (Gjerløw 1977).

24

1.1. Norwegian fragment studies

Most of the findings are apparently fragments of Latin books of ascetic contents and therefore not of particular interest for us, apart from that they in general are beautifully written and seem ancient.

The liturgical fragments were not used as witnesses to the earliest phases of Nordic book culture, due to their religious contents in combination with their Latin language. Liturgical texts were not historical texts and therefore were not considered primary sources for the study of Norwegian history or language, subjects of major interest in the ninteenth century. Latin fragments which were considered to shed light on Norwegian history or local Norwegian saints were published at an early date by the historian Gustav Storm (1845-1903) in Monumenta historica Norvegiae (Storm 1880), but other Latin fragments were left unstudied. In 1910 S. A. Sørensen tried to draw attention to the importance of the Latin fragments from liturgical, religious and philosophical works, from lawbooks and medical books, and expressed the wish that these Latin texts be studied to increase the knowledge of medieval life in Norway (Sørensen 1910, 46). With the exception of some of the liturgical fragments, he did not succeed, as the non-liturgical parts of the Latin material have been left relatively unstudied until now. The liturgical material, on the other hand, came to interest musicologists when Georg Reiss in 1906 discovered the liturgy for St. Olav, the Norwegian patron saint, and chose this as the topic for his dissertation (Reiss 1912). Still, the liturgical fragments were basically valued as the carriers of medieval music, not as witnesses to medieval books.

1.1.2. Liturgy as law The scholar Oluf Kolsrud (1885-1945) separated the fragments into envelopes and numbered them in the system which up to now has been the governing one (cf. Pettersen 2003, 52). In 1920 he suggested in a letter to the Commission for the edition of the Old Laws of Norway (Kommisjonen til utgivelse av Norges Gamle Love) the use of the liturgical fragments in the NRA to make editions of the liturgical books of Nidaros as a side-branch of the work to edit the Old laws of Norway. He argued that the liturgical books had the character of law and belonged in a complete edition of 25

Part I: To study manuscript fragments the Old laws of Norway as an addition to the ecclesiastical section. He also claimed that there was material for the following books: 1. Missale, 2. Breviarium, 3. Ordinarius, 4. Graduale, 5. Antiphonarium, 6. Kalendarium, 7. Manuale. Although positive in principle, the reply was that the work on the Old Laws should come closer to completion before venturing into the field of liturgy (Fæhn 1962, vii). A new initiative from Kolsrud in 1939 was obstructed by the war, and, finally, by his death in 1945. In 1947 Erik Eggen (1877-1957) was asked to undertake a registration of the fragments in the NRA, which he set out to do the following year (Eggen 1968, xi). While performing this task he became interested in their contents, and focused on fragments with sequences in particular, as this was a topic which had interested him for some time (ibid., xii). Although he continued working with the registration and the description of the contents of the envelopes, he did not come close to finishing this work. In 1954 Helge Fæhn reintroduced Kolsrud’s plans of liturgical editions, and received governmental funding for the first work in the series “Libri liturgici provinciae Nidrosiensis medii aevi”, namely Manuale Norvegicum (Presta handbók) (Fæhn 1962). The two following and, for the present, last books in the “Libri liturgici” series were edited by Lilli Gjerløw (1910-98). Gjerløw was employed at Norsk Historisk Kjeldeskrift-Institutt (Norwegian Institute of Historical Documents), but, as her research to a large degree depended on the NRA holdings, her work was done in close collaboration with this institution (cf. Pettersen 2003, 53). In 1968 the Ordo Nidrosiensis Ecclesiae (Orðubók) came out, and in 1979 the Antiphonarium Nidrosiensis Ecclesiae. In addition to antiphoner fragments, Gjerløw also studied closely a group of missal fragments, and published studies of the oldest missals (Gjerløw 1961) and of missals used in the bishoprics of Bergen and Oslo in the Middle Ages (Gjerløw 1970, 1974). Gjerløw is the giant whose shoulders we stand on; her work and publications on the Nidaros liturgy are invaluable, and quite unique in a Nordic context.

26

1.1. Norwegian fragment studies 1.1.3. New initiative The last chapter in fragment research sprang out of an initiative to register NRA fragments in a digital database. This relatively recent initiative came from the musicologists, more specifically from Andreas Haug, then Professor at the Centre for medieval studies (SFM) at the University in Trondheim (NTNU). After meetings in 2000 and 2001 an application was sent to the Norwegian Research Council and 22

received funding. The project to register liturgical fragments in a database started in 2002 as a collaboration between SFM and the NRA.23 The task of doing the registration was given to Espen Karlsen, and in the period between 2002 and 2006 a considerable number of fragments were entered into the database, starting with the missals and graduals. Karlsen used Gjerløw’s catalogue as a point of departure, but made additions both in terms of identified missals and graduals, and in terms of single fragments supplementing already identified manuscripts. When the last period of funding from the Resarch Council ended in 2006, the work on the database was taken over by Gunnar I. Pettersen.24 Although the process of registering the fragments in the NRA has made substantial progress during the last few years, such a task is time-consuming and it is not yet finished. The challenges involved in studying manuscript fragments in Norway are considerable, as the fragments are often small, detached from their original context and difficult to deal with without expert knowledge. A relatively large number of fragments have not yet received close study by either Eggen, Gjerløw or Karlsen and have therefore not been identified or dated. In connection with the NRA database project important results have come after the three international workshops. The work done so far in the NRA workshops reaches far beyond being a tool for musicological studies or a digital catalogue. Emerging from this material are the contours of • the introduction of the Latin alphabet 22

Cf. written summary of 04.04.01 by Gunnar I. Pettersen from the meeting in Oslo 19.02.01 between NRA staff, musicologists and participants in the Swedish MPO-project. 23 The Oslo project has used the same type of database as the one used in the Swedish National Archives in Stockholm. 24 I thank Espen Karlsen and Gunnar I. Pettersen for the information about the continued work with the database.

27

Part I: To study manuscript fragments • the introduction of Christianity • the introduction of the Latin language • the establishment of the first Norwegian scribal communities The first results from these workshops and from Espen Karlsen’s work can be found in recent publications from Karlsen (2003, 2005). A workshop was held at the initiative of the Centre for Medieval Studies (CMS) in Bergen in October 2005, stressing the importance of Nordic and international collaboration, as well as the value of interdisciplinarity in the field of fragment research.25 A brief presentation of the current status of fragment cataloguing and research was given from each Nordic country (cf. the workshop report, Ommundsen 2006a). In some cases, as for the Swedish National Archives, the situation was dramatically improved from the last Nordic conference on fragment research held in Stockholm in 1993, as the project to enter all the 22 700 fragments in the Swedish National Archives in Stockholm into a database is now complete (cf. Brunius 1994, 2005). Broader studies involving fragments, like this one, are a natural consequence of the initiatives made since the early 1990’s. Scholarship and the efforts to facilitate searches in the fragment collections depend mutually on each other.

1.1.4. Fiefs and provenance A final note on the past fragment research is connected to issues regarding their reuse. How were the fragments collected, where and by whom were they dismantled and used to bind accounts? To what extent is the fief on the account relevant to the medieval provenance of the manuscript in the binding? Views regarding this have been that either the bailiffs bound accounts locally in the smaller fiefs (Munch 1847), or they were bound centrally in the larger fiefs, in the castles in Oslo, Trondheim, Bergen and Stavanger (Storm 1893). These different views have recently been discussed by Pettersen (2003, 55-8). As there was probably not one single practice, each case should be considered separately. Although fragments from a manuscript

25

Workshop material was distributed in advance (Ommundsen 2005).

28

1.1. Norwegian fragment studies can sometimes be placed within a region with some degree of confidence, one can only very rarely say exactly where a liturgical manuscript was used. That the practice was not regular does not mean, however, that the regions indicated on the accounts cannot be of any use to us when looking for a secondary provenance. There are at least two indications that a manuscript was dismantled and used to bind accounts locally. If pieces from one manuscript were used in one local centre over a period of several years and not found in bindings from any other centres, this can 26

indicate a local binding. A similar case occurs if pieces from one manuscript were used to bind accounts for several smaller fiefs, all connected to one region or castle, over a period of time. Then it is likely that the manuscript was dismantled and used for a binding in that region’s centre or castle. Another type of indicator for a secondary provenance is if pieces of one manuscript are found in one region’s accounts in the NRA and in the local archives or libraries of that same region. Examples of this type occur in Stavanger (Seqv 38 add) and Trondheim (Seqv 31 add). There are a few existing studies of the specific fiefs and their organisation, which, although they are quite old, are still quite useful, especially Danmark-Norges Len og Lensmænd 1596-1660 (Erslev 1885) and Lensprincipet i Norden (Lie 1907), particularly on the fiefs in Norway 1536-1660, pp. 70-102. Here, for the help of the reader, there will merely be supplied a map of the fiefs and some of the centres for reference. Regarding the fragments from the liturgical manuscripts I will evaluate in each case the most likely relationship between the manuscript/fragment and the fief concerned in the account.

26

As suggested by Gunnar I. Pettersen at the workshop in March/April 2005.

29

Part I: To study manuscript fragments Map 1: Map of the Norwegian fiefs c. 1640. The map is made using map 71 in the Historical atlas of Norgeshistorie (Cappelen 1980).

30

1.2. The palaeographical traditions

1.2. The palaeographical traditions Palaeography is one of the most important tools in fragment research. What has been most surprising in the NRA workshops is the potential palaeography has proven to have to argue for or against the origin of a given manuscript. This does not only apply to the identification of the locally produced manuscripts, but also for recognising manuscripts from different European regions. While this aspect will be treated in more detail in chapter 3, the following is devoted to other functions of palaeography in international scholarship, on the relationship between Old Norse and Latin palaeography in Norway, and on the importance of codicology in fragment research.

1.2.1. The purposes of palaeography The birth of palaeography was closely linked to diplomatics and the study of documents, particularly with a concern to reveal forged documents.27 This function is now marginal.28 Palaeography soon became an important auxiliary tool for both philology and history. Whether connected to diplomatics, philological studies or history, one fundamental function of palaeography has been the proper understanding and transcription of medieval texts, or, as it is formulated in the English

translation

of

Bernhard

Bishoff’s

Latin Palaeography,

”the

correct

decipherment of old handwriting” (Bischoff 1990, 2, German orig. 1979). A third important function is to provide guidelines for the proper dating of manuscripts. There are challenges connected to the assignment of dates, and unless there is some specific internal evidence in the manuscript, it is difficult to narrow the date down more than to a period equivalent to a generation or a lifetime of writing (cf. for instance Karlsson 1999, 146).

27

This was already a concern for pope Innocent III (1198-1216) who issued a decretal on the detection of false bullae (Boyle 1992). For the seventeenth century discussion for the criteria for the “discrimen veri et falsi” between Daniel van Papenbroeck in Acta sanctorum (1675) and Jean Mabillon in De re diplomatica (1681), see ibid, 83-4. 28 Bernhard Bischoff treats the forging of manuscripts, which occurs more rarely than that of documents, in a brief appendix in Latin Palaeography (Bischoff 1990, 46-7).

31

Part I: To study manuscript fragments A fourth function is the assignment of origin, which is and will remain the biggest challenge. To be able to indicate both date and origin is of great importance for someone involved in the work of making a catalogue of manuscripts. Still, dating is treated more thoroughly than indicating a regional origin for a manuscript without internal information regarding its making or former whereabouts. The use of palaeography to localise manuscripts will be discussed below. For now this aspect of palaeography will merely be adressed by a reference to the cataloguer Gerhardt Powitz. In his article “Datieren und lokalisieren nach der Schrift” he asks to what extent stylistical peculiarites in script can be decribed and taught through text books at all, and suggests the following conclusion: Kommt es nicht viel mehr darauf an, im Umgang mit einer möglichst grossen Anzahl von Handschriften den Blick zu schärfen, um dann – gegründet auf Erfahrung und die Sicherheit des geschulten Auges – ein eher persönliches, intuitives Urteil zu wagen? (Powitz 1976, 136)

As soon as one turns from dating to the determining of origin, the idea of the “eye” developing after being exposed to a large number of manuscripts seems to be the general perception. And who can argue against the value of experience? Powitz fortunately adds a final comment, which is fundamental in all fields based on seeing and observing: “das Auge muss sehen lernen” (Powitz 1976, 136). An important aspect of palaeography, especially when searching for the origin of a scribe is to know which features to look for. The more we know, the more we see, and the art of “looking” seems to be something which can be trained and taught. But the eye needs a checklist as it learns how to see. Such a checklist will be provided in chapter 3.

1.2.2. Derolez and Scandinavia An attempt to make “das Auge sehen lernen” was the admirable goal of Albert Derolez, who published his book The Palaeography of Gothic Manuscript Books in 2003. An important aspect of Derolez’s book was that he questions the notion that palaeography cannot be taught (Derolez 2003, 1-2). Without the hope of acquiring knowledge through books and teachers, the road to achieving “the eye” for 32

1.2. The palaeographical traditions palaeography would simply seem to long and strenuous for many scholars, particularly for those far away from the large manuscript collections of the central parts of Europe. Derolez’s book has had a stronger impact in Scandinavia than most other books on Latin palaeography, first and foremost since its main focus was the period in which most books were being written (and still remain to be studied) in Scandinavia, namely after 1200. Much of the former palaeographical literature, especially on the Carolingian book hands, may be interesting as an introduction to the history of writing, but with hardly any practical consequences to Scandianavians working with locally produced manuscripts from the thirteenth or fourteenth century. The pedagogical value of Derolez’ approach is considerable. The many plates and the simple nomenclature29 makes it possible to use in practical “learning” as well as teaching, and balance the categories and periods required for teaching and searchable digital catalogues and the need for “recognizing the uniqueness of every manuscript”, as expressed by Bischoff (1990, 3). Although Derolez emphasises the importance of the Catalogue of dated manuscripts series, this blessing has not often reached the libraries in the northern periphery. If a book on palaeography is much larger than Derolez’s or S. Harrison Thomson’s Latin Bookhands of the later Middle Ages 1100-1500 (1969), chances are that it is unavailable in a Norwegian library, perhaps with the exception of the University Library in Oslo. The immediate availability of Derolez’s book is one of the main reasons for its impact and influence. After Derolez’s book was published initiatives were made by Old Norse philologists, first by Guðvarður Már Gunnlaugsson in Iceland, and then by Odd Einar Haugen in Norway, to “Europeanize” the terminology of the Old Norse palaeography. In his

29

A book like Michelle Brown’s A Guide to Western Historical Scripts from Antiquity to 1600 (Brown 1990) can be useful, but never held the same potential for influence and use in Scandinavia for two primary reasons. One is the limited pedagogical value of descriptions like littera minuscula protogothica textualis libraria media/formata (cf. Brown 1990, 74). The other is the perspective of European script as “Continental” vs. “English” (although this in many cases is the perspective of this current study as well). Still, to contrast “northern” with “southern” as in the case of Derolez seems more suitable for a general introduction to European scripts.

33

Part I: To study manuscript fragments latest introduction to Old Norse philology Haugen introduces the term “Pregothic” (“førgotisk”) for thirteenth century script with a reference to Derolez (Haugen 2004, 197), instead of the previous term “Younger Carolingian-insular script”.

1.2.3. Old Norse and Latin palaeography in Norway In Norway Old Norse palaeography and Latin palaeography have existed side by side for at least half a century. What is here referred to as “Old Norse palaeography” is basically a subbranch of Latin palaeography which has formed its own tradition and terminology. While Old Norse palaeography was thriving for a large part of the twentieth century and was regularly applied to the study of Old Norse texts, Latin palaeography was only kept alive by Lilli Gjerløw for her research on liturgical fragments. While Gjerløw’s efforts had little impact on Norwegian palaeography in general (which can only serve as a reminder of the importance of interdisciplinary collaboration), contact between the Latin and the Old Norse philologists has led to promising results since then.30 Attempts have been made to narrow the distance between the Old Norse and the Latin palaeography as a natural consequence of the fact that the book and scribal culture – and the scribes themselves – are basically the same. There are two principal features dividing the Old Norse palaeography from the Latin. The first is the method of dating, which in Latin script is based on features in the script, while it in Old Norse is a combination of palaeographical and linguistic indicators.31 Another barrier between the Old Norse and Latin palaeography has been the terminology, particularly regarding script and periodisation. The Norwegian linguist and palaeographer Didrik Arup Seip divided Norwegian script into three periods (without naming them): Before 1225, 1225-1300 and after 1300 (Seip 1954). The Swedish palaeographer Lars Svensson named the period before

30

Lars Boje Mortensen (Latin) and Odd Einar Haugen (Old Norse) in Bergen have collaborated with interdisciplinary student seminars in palaeography, like “Book and script in the Middle Ages” during the spring of 1998. The new developments in Norwegian palaeography also include both branches. 31 I thank Odd Einar Haugen for pointing out this difference.

34

1.2. The palaeographical traditions 1300 Older and Younger Carolingian-Insular and the time after 1300 Gothic (Svensson 1974), a terminology taken up by Norwegian scholars. The corresponding terms in Latin palaeography would be Pregothic or Protogothic for the century before 1200 and Gothic for the centuries after 1200. The unfortunate effect of two such different systems is that the same Norwegian scribe writing in the twelfth century would, according to Old Norse terminology, be writing an Older Carolingian Insular script, while in Latin a Pregothic or Protogothic script. Two products of the same Norwegian scribe writing one century later, in the thirteenth century, would, in the Old Norse tradition, be said to write a Younger Carolingian Insular, while his script in Latin would be described as Gothic. As mentioned above, after Derolez “Pregothic” or “førgotisk” was introduced as a term in Old Norse, but here for the script of the thirteenth century. Although the adoption of the term Pregothic made the terminology of the two traditions more similar, they still do not correspond completely, as illustrated in the following table: Table 1: Periodisations and terminology in Latin and Old Norse palaeography Latin palaeography

780-1100

Carolingian script

c. 1100-1200

Pregothic

1200-1300

(cf.

Derolez

Old Norse terminology

Old Norse terminology

ante-Derolez

post-Derolez

(Haugen 2002, 829-30)

(Haugen 2004, 202)

Older

Carolingian-Insular

Carolingian-

2003) or Protogothic (cf.

Insular script (up to c.

Brown 1990) script

1225)

Gothic script

Younger

Carolingian-

script Pregothic script

Insular script (c. 12251300) 1300 onwards

Gothic

textual

and

Gothic script

cursive script (from c. 1300)

One obvious question presents itself from this table: was “Pregothic” written in the western parts of Scandinavia a century later compared to the rest of Europe? Although we may say that there was a certain delay before the “new” style of writing 35

Part I: To study manuscript fragments took hold in the North, it seems that a “reasonable delay” would be the 25 years registered by Seip when he set a “shift” in style of writing at approximately 1225. It is true, however, that several thirteenth century Scandinavian manuscripts display a general roundness or ovality, and not the angularity often connected with “Gothic”. Dated Norwegian documents also display large individual variations among the thirteenth century scribes regarding the use of biting (cf. Hødnebø 1960). Still, the cause of the different use of the terms seems to be different opinions regarding how far the “Gothisation” of script should have come before writing can be called “Gothic”. Old Norse palaeography traditionally emphasises the verticality and angularity of the Gothic to a larger degree, and sees the first inconsistent signs of biting or round r (as pointed out in the rule of Meyer) as signs of Pregothic rather than Gothic.32 I have followed the practice of Latin palaeography in that if something contains biting to some degree and is datable to the thirteenth century, I refer to it as “Gothic”. Periodisations and “labels” on script may seem unnecessary when the focus of our study is specific manuscripts written by scribes who naturally did not know how their products would be assessed, evaluated and “labelled”. Still, to develop means to explain and simplify complex material, either for pedagogical purposes or for making something digitally searchable, can be useful. These types of distinctions will always represent a first step – either at the level of research or the level of knowledge of the individual student or scholar, and should only be used as a tool, not a rule.33 Work done in the field of Old Norse palaeography is important and useful also for the study of Latin fragments. Kr. Kålund published his Palæografisk Atlas with plates (1903-07). Didrik Arup Seip (1884-1963) is still regarded as a great authority for his study on the palaeography of Norwegian and Icelandic manuscripts (Seip 1954) and his Norsk språkhistorie til 1370 (Seip 1955). In his Palæografi Seip identified fourteen 32

Lars Svensson mentions round r after other round letters than o as a feature in the “Younger Carolingian Insular” from the middle of the thirteenth century (Svensson 1974, 203). Although Odd Einar Haugen recognises biting or fusion as a Gothic feature, he interprets the gradual introduction of these features in the thirteenth century as a Pregothic characteristic (Haugen 2004, 188), not as the first stages in the Gothic script. 33 cf. workshop discussion in Ommundsen 2006a.

36

1.2. The palaeographical traditions hands divided between eight manuscripts for the period before c. 1225, and placed the first four manuscripts (with four hands) in Nidaros and the last four manuscripts (with ten hands) in Bergen. For the period between 1225 and 1300 he listed 27 of the most important Norwegian manuscripts, with information on scribes where possible. Seip also gave the palaeographic analysis of scribes writing in Latin, in Manuale Norvegicum (Fæhn 1962), with several hands from the thirteenth and fourteenth century. The importance of bringing Old Norse and Latin palaeography together was also recognised by Lilli Gjerløw. In addition to consulting European authorities like N. R. Ker, she also related and referred to Norwegian scholars like Seip. After identifying missals she thought were probably written in Norway Gjerløw concluded that the missals could be added to Seip’s list of sources written by Norwegian scribes before 1225 (Gjerløw 1970, 112). In her edition of the Nidaros ordinal she also treated four Norwegian scribes or scribal centres, three of which were writing in both Old Norse and Latin (Gjerløw 1968, 34-8). Scribes are not only found in books, but also in documents. During the last decades of his life Eivind Vågslid made a courageous attempt to distinguish and identify the different scribes and notaries in Norway. He studied all Norwegian documents, both in Latin and the Old Norse vernacular, from 1175-1400 (c. 3650 letters), and presented 800 named scribes, as well as a large number of unidentified hands. He also made an effort to divide these hands into scribal schools (Vågslid 1989). He has been critisised from two sides. On the one hand he is critisised for identifying too few scribes, on the other for identifying too many.34 This problem will be further discussed in chapter 3.3. on the grouping of manuscripts. As the work with the Latin fragment continues, it is likely that more scribes writing in both Latin and the vernacular will be identified. Future cooperation will therefore be beneficial for both palaeographical traditions, regardless of terminology.

34

Cf. Haugen 2004, 213, who refers to the objections regarding too few hands one side (Hagland 1990), and too many on the other (Blom 1992; Bakken 1997).

37

Part I: To study manuscript fragments 1.2.4. Codicology without the codex As palaeography deals with the script, codicology deals with the other aspects of the book, such as the material, the binding, the lay-out of the page and the decoration, including the initials. How do we deal with “book study” when a book no longer exists, and all that is left of it is one or more pieces of parchment? The reality is that codicology becomes almost more important when a book is destroyed, and fragments are the only remains of it: When so little remains it is necessary to evaluate every piece of evidence available to us. The “art of measurement”, which made Bischoff almost audibly sigh through the pages (cf. Bishoff 1990, 1; Derolez 2003, 7-8), is crucial when arguing for or against the link between a fragment and an identified once-complete manuscript. We need to measure that which is measurable to avoid making mistakes when joining fragments together in our “phantom-codices”, “reconstructed manuscripts” or “once-complete manuscripts”, or whatever we decide to call the few remains of the objects of our interest. Derolez refers to palaeography as a subdiscipline of codicology (2003, 10). A section devoted to codicology opens Bishcoff’s classic Latin Palaeography (1990, German orig. 1979). The term codicology was introduced by Masai (1950) in an attempt to separate palaeography from the other aspects of the medieval book. J. P. Gumbert has given a useful introduction to the field in his article “Fifty years of Codicology” in Archiv für Diplomatik (2004). He emphasises that codicology in general and in its original intention is connected to the study of manuscripts as bearers of texts. Still he recognises the study of the book as a material object and a result of craftmanship, for which he suggests the term “material codicology” (Gumbert 2004, 507). In codicology as in palaeography the ideal is to develop an “eye”, as opposed to the focus on a set of details. Still, Gumbert also acknowledges the need to specify that which is seen: “The codicological eye is a way of looking, not a sum of facts. Yet a sum of facts will be helpful” (Gumbert 2004, 511). Although many advances in codicology have been made during the last fifty years, here I will only mention a few fields with particular use for this study.

38

1.2. The palaeographical traditions When it comes to the terminology for codicological descriptions Denis Muzerelle’s Vocabulaire codicologique (1985), and particularly the online version also giving the English,

German

and

Italian

equivalents

to

the

French

terms,

(http://vocabulaire.irht.cnrs.fr), is most useful. Regarding parchment the most complete introduction is the book called Pergament, edited by Peter Rück (1991). That parchment really has a potential in the studies of fragment and for the identification of locally made scriptoria is indicated by Michael Gullick, who has noticed that the parchment of Swedish documents, and books believed to be locally made, have a particular “nappy” quality, making the hair and flesh side appear relatively similar (Gullick 2005, 59). The same quality can to a certain extent be found in Norwegian local parchment. In a few lucky cases so much of the margins remain that the slits or holes from the bindings are visible, which makes the study of binding techniques relevant even when a book has been dismantled. Several fragments have some form of initials, either simple or decorated. Most initials are not decorated enough to become interesting for the art historians, and the study of these generally get put into “codicology”. For the study of initials and other decorations there is help in the works of Otto Mazal on Romanesque and Gothic book art (1975; 1978), Jonathan J. G. Alexander on the decorated letter (1978a), and his closer study on the twelfth century English arabesque initial (1978b), not to mention his Medieval illuminators and their methods of work (1992). Francois Avril and Patricia Stirnemann have edited a very useful volume with many plates on the insular illuminated manuscript in the Bibliothèque Nationale (Avril and Stirnemann 1987) and Patricia Stirnemann has done important work on the penflourished initial (1990). Michael Gullick has exemplified the benefits of studying the initials, in his work on the Swedish fragments (Gullick 2005). The possibility to use modern techniques in the study of parchment and pigments is there. Yet so much work remains to be done with the Norwegian material that we do

39

Part I: To study manuscript fragments not quite know which questions to ask yet. A more traditional approach, with looking and sorting through the material should come first. One feature of liturgical manuscripts which belongs in the chapter of codicology is the musical notation. Musical notation, either in the form of neumes in campo aperto or staff notation is a field of its own, and also an important source for additional information about a given fragment. Bruno Stäblein’s Schriftbild der einstimmigen Musik (1975), providing an overview of different types of neumatic notation and the introduction of square notation or Quadratschrift, is a standard work on musical notation. So is Madeleine Bernard’s Répertoire de manuscrits médiévaux contenant des notations musicales (Bernard 1965, 1966, 1974). John Haines has pointed to a certain parallell between palaeography of script and paléographie musicale regarding the overweight of study on the neumes of the earlier period compared to the later, Gothic square notation, and expresses the need for a nomenclature and graphic history also for square notation (Haines 2004). Musical palaeography requires the development of yet another “eye”. It can be an advantage to consult musicologists with experience, who can often extract impressive information from looking at a notated manuscript page. This means that although the selection of a liturgical genre, like the sequence in this case, is more challenging, there is also one more source of information. From this paragraph on musical notation, it is time to move on to the music itself, and take a closer look at the sequence, and the most important research done in this field.

40

1.3. The sequence: rise, fall and resurrection

1.3. The sequence: rise, fall and resurrection Although this is not a musicological study, it is still a contribution to the field of sequence research. While musicologists can extract information from a particular repertory or the specific textual or melodic variant of a sequence, the study of the manuscript may shed additional light on sequences and their transmission. And palaeographers in turn benefit from what has already been discovered in almost two centuries of sequence studies. The sequence is a type of medieval chant used to highlight particular feasts of the church year, and sequence texts often emphasise joy, song and celebration. The history of the sequence can basically be summarized as one of a rapidly growing popularity from the ninth century onwards, and a marked decline in the fifteenth century, with the near abolition of sequences in the sixteenth. During this time, several thousand sequences were composed in various regions of Europe, both for pan-European ecclesiastical celebrations and for local saints and feasts. In the last two centuries sequence studies have been an important field of research.

1.3.1. The sequence – history, form and contents The sequence was developed during the Carolingian age and was sung after the Alleluia in the celebration of Mass. In other words, the sequence was a “late” liturgical element originating north of the Alps, and not part of the original Roman rite. Its use was tolerated, but not prescribed by the official Roman church. This left a great deal of freedom regarding the sequences to the later makers of local rites. The first named composer of sequence texts was Notker Balbulus of Sankt Gallen (c. 840912) with his Liber ymnorum (c. 880).35 By the time the sequence was introduced to Scandinavia with Christianity the genre was already established, and the problems regarding its origin will not be discussed here.36 35

The Notker repertory (forty sequence texts to thirty-two melodies) was established by Wolfram von den Steinen in his Notker der Dichter und seine geistige Welt (Steinen 1948). 36 For a short introduction to the origin and first medieval references to the sequence, see for instance Western Plainchant (Hiley 1993), 185-189, or Lori Kruckenberg’s article in Die Musik in Geschichte und Gegenwart (Kruckenberg 1998) .

41

Part I: To study manuscript fragments

One sequence melody could often be used for a large number of texts, which were written with particular saints or feast days in mind. The lyrics follow the melody syllabically, which is a very recognisable feature on an annotated page, compared to the melismatic movements of the antiphonal or responsorial chants. As it was so closely connected to the Alleluia, the sequence was centrally placed in the mass, between the Epistle and Gospel readings (see table below). Although sequence texts can refer to Biblical or hagiographical texts in an inter-textual relationship, they represent their own textual entity, and present their own version of a saint’s life and miracles. The text of a sequence often points to the miracle in a kind of internal code, with references rather than explanations, and it is presupposed that the listener (and singer) knows the contents from other sources. The most important formal characteristic of a sequence of the older type up to c. 1100, apart from the mentioned syllabical text-music relationship, is the organisation of the strophes into verse pairs. In the older type, the strophes not only differ with respect to melody, but also in the size and structure of the text. While the melodic lines and textual structure change from one strophe to the next, the verse pairs of each strophe have the same melody and form. The pairs of equal melodic and textual form, different for each strophe, separate the sequence from a hymn, where each strophe is sung to the same melody. In numerous sequences for a saint, the opening strophe encourages the people to rejoice and sing in celebration. The word “hodie” often occurs in sequences, either in the opening or in reference to the events leading to the martyrdom of the saint (on the date of the celebration). In this manner the sequence connects the saint to the present, actualising the events of the past. The opening is followed by several strophes with elements from the saint’s vita, his or her miracles and virtues, and finally a request for the saint’s intervention. The three main elements can be listed as follows: 1) Let us today celebrate the feast of saint NN. 2) These are the miracles and virtues of saint NN.

42

1.3. The sequence: rise, fall and resurrection 3) May saint NN protect us and ensure our salvation. As the sequence often is a rendering, although in poetic form, of a saint’s life, it is a way of putting the events of a saint’s life right between the readings from the Bible in the mass, giving the saint added authority. And as the sequence (or rather the cantor or choir on behalf of the congregation) turns to the saint for help, the sequence also becomes a kind of prayer, connected to the proper prayers of the Mass celebration.

The medieval Mass Main elements: Lessons (epistle and gospel) and communion Main books: Missal (all elements) or gradual (songs only) Introitus Kyrie eleison Gloria in excelsis Oratio collecta Epistola (first reading) Graduale (responsorial song) Alleluia (with verse) Sequentia Evangelium (second reading) Credo Offertorium Praefatio (”vere dignum”) Sanctus – Benedictus Canon (”te igitur”) Oratio super oblata (or secreta) Pater noster Agnus Dei Communio (the distribution of bread and wine) Communio (antiphonal song) Oratio ad complendum (or postcommunio) Ite missa est/Benedicamus

The patron saint of Nidaros, St. Olav (d. 1030), had a proper sequence, Lux illuxit letabunda, written in his honour. Although no fragments from pre-1200 manuscripts contain the sequence, it was most likely written in the late twelfth century (cf. Reiss 1912, 17). At the mass for St. Olav on 29 July the cantor or choir would sing in honour

43

Part I: To study manuscript fragments of the joyous and glorious light shining on the land of darkness, imploring for his protection from the evils of this world in the final strophe: 8a. Rex et martir triumphalis tutor noster specialis tua proles spiritalis sit ab huius mundi malis te tuente libera. 8b. Quos infestat vis carnalis corruptela generalis pestis potens et letalis nos sub tuis tutos alis tua salvet dextera. 8a. Triumphant king and martyr, our special protector, free your spiritual offspring from the evils of this world with your protection. 8b. As we are threatened by the forces of the flesh, the general depravity, the powerful and lethal pestilence, may your right hand place us safely underneath your wings.

While the most important saints, like St. Olav, would get one or more proper 37

sequences, others did not, but were assigned sequences from the commons of saints. These sequences could be applicable for several saints, and in a few of them the insertion of a name is necessary. The use of “NN” can actually occur in a written text to be supplied with the proper name for the performance. One example is the sequence Laudes debitas deo (AH 54, no. 62) for a virgin martyr, in the Nidaros ordinal prescribed for St. Agatha, in Missale Nidrosiense (1519) with the rubric De una virgine:38 2a. Et hodie cum sanctis suis coronavit virginem inclitam 2b. In polorum sedibus per palmam martirii NN.

37

39

Other later sequences exist for St. Olav, like Postquam calix Babylonis, Predicasti dei care and Veneremur sanctum istum. 38 In spite of the general rubric in Missale Nidrosiense, and the “NN” in the v. 2b, the name of Sta. Lucia occurs in v. 5b. For this reason the sequence in AH 54 (no. 62) is published with the rubric De sancta Lucia. Missale Nidrosiense is listed as its only source. 39 For this sequence, see Missale Nidrosiense 1519, 596, Eggen 1968, 307-9, Gjerløw 1968, 433. The NN is to be understood as an accusative (and not a genitive), as the sequence has versepair-by-versepair end-rhyme throughout. The end-rhyme indicates that this is not a sequence of such an early kind as Eggen suggests in the discussion of its possible Norse origin (Eggen 1968, 309).

44

1.3. The sequence: rise, fall and resurrection 2a. And today with his saints he crowned the renowned virgin NN 2b. in the heavenly quarters as she won the victory of martyrdom.

Common sequences were applied for some of the saints in the Nidaros ordinal (introduced right after 1200), like the local saints St. Hallvard (15 May) and Sancti in Selio (8 July) (cf. Gjerløw 1968). The fortunate discovery of a proper St. Hallvard’s sequence on an Icelandic manuscript fragment revealed that St. Hallvard, presumably after the introduction of the ordinal, was given a proper sequence, Lux illuxit... lux est nobis (Reiss 1912, 44; Eggen 1968, 184). No later sequence has (yet) been discovered for the saints of Selja or Sancta Sunniva, which does not, of course, mean that one never existed.40 After the immense popularity of the sequence in the twelfth and thirteenth century, the first voices were raised against sequences in the fourteenth century (cf. Kruckenberg-Goldenstein 1997, 13). In the sixteenth century the epoque of the sequence was over, as the Council of Trent (1545-1563) reduced the number of “approved” sequences in the official Roman liturgy to five.

1.3.2. International sequence research Modern sequence research began in the middle of the nineteenth century, with the publication of two volumes of the Thesaurus hymnologicus dedicated to sequences.41 The sequences were divided into two main groups on stylistic grounds. The first group, from the ninth to the eleventh century, was characterised by an elevated prose style, referred to as Kunstprosa, with texts formed to underline the melodic lines. In

40

No liturgical fragments have as of yet been discovered containing any liturgy for the feast of the saints of Selja. This is unfortunate, as the inclusion of a proper Office (without notation) for the saints of Selja in the printed Nidaros breviary (Breviarium Nidrosiense 1519) shows that proper liturgy for the feast did exist. The presence of the texts for the Office in the printed breviary provides an opportunity for scholars to study the texts, but unfortunately no fragment containing the music has been found. Although there are no indications that a sequence for the Selja saints ever existed, there seems to have been the required skill for sequence composition in Bergen in the thirteenth century, as shown by the weddingsong Ex te lux oritur, written in the form of a sequence for a royal wedding in 1281 (Kolsrud and Reiss 1913). 41 The second and fifth volume appearing in 1841 and 1855 were devoted to sequences (Daniel 18411856).

45

Part I: To study manuscript fragments other words, in the earliest phase the melody was the primary element, and the text secondary. The second group followed a pattern of rhythm and rhyme. Another work with a major impact on the study of sequences was the Analecta hymnica (Analecta hymnica medii aevi (AH)) edited by C. Blume, G. Dreves and H. M. Bannister in 55 volumes in 1886-1922. Its purpose was to edit a multitude of liturgical genres, among them sequences, hymns, tropes and rhymed offices, and it printed more than 3000 sequence texts for the first time. In addition, it also re-edited the sequences of the Thesaurus hymnologicus, making a total number of c. 4500 sequence texts (Blume 1922, v). Blume and Bannister pointed out that stylistically there was a third group of sequences, placed between the “first epoch” and “second epoch”, and subsequently called these sequences “Sequentiae transitoriae” or “Prosen ‘des Übergangsstiles’” (Blume and Bannister 1915, v). Although Blume and Bannister insisted that the “transitional sequences” were not equal to a timeframe or Zeitperiode (Blume and Bannister 1915, vi), it is possible to advance the following paradigm with estimations of dates (cf. Kruckenberg-Goldenstein 1997, 8): 42 1. First epoch (before 1050) - Kunstprosa 2. Transitional sequences (c. 1050-1150) – diversity and experimentation with rhythm and rhyme 3. Second epoch (after 1150) – consistent rhythm and rhyme Regarding the earliest phase of the sequence, Blume and Bannister found that Europe was divided into two regions or spheres, the Germano-Italienische Kreis (Germany, the Netherlands and Italy) and the Gallo-anglicanische Kreis (England, France and Spain). Within the countries of the same sphere there was a frequent exchange of sequences, but seldom between the different spheres. Only a limited number of sequences crossed borders, and, if so, it would be between France and Italy or between England and Germany, but seldom between France and Germany (Blume and Bannister 1911, xxix-xxx). The position of Scandinavia was not considered due to the lack of pre-1100 sources. 42

Kruckenberg discusses the problem with dates in Kruckeberg-Goldenstein 1997, 140-6.

46

1.3. The sequence: rise, fall and resurrection

The main focus of scholars has primarily been different aspects of the early sequence (Crocker 1977) – the Notker-tradition (Steinen 1948), about the relationship between the sequence and the Alleluia of the mass (Bower 2003) or the relationship between text and melody and the written representation of text and music (Haug 1991, 1987, 1991). Margot Fassler has worked with the late sequences of Paris and St. Victor (Fassler 1993), and Lori Kruckenberg has studied the sequence repertory of the southern German abbey of Hirsau (Kruckenberg 1999). Lori Kruckenberg has also done important work on the type of sequence referred to by Blume and Bannister as “transitional sequences”, now labelled sequentiae novae (Kruckenberg 2006a, 36). Among other things, she addressed the issue of European reception and found no Rezeptionsbarriere similar to that of the first epoch sequences. She found that unlike the first and second epoch sequences, the sequences in the “transitional style” did not form unified repertories, and that they appeared in a new type of source occurring after c. 1050 which, to a certain degree, included both eastern and western sequences (Kruckenberg-Goldenstein 1997, 159-63). The remarkable size and eclectic nature of the Nidaros ordinal has made the Nidaros sequence repertory interesting for international study, as sequences from both the German and the Anglo-French tradition and from all epochs were selected for use in Nidaros (Gjerløw 1968, 433). The recently published results of the Nidaros sequence project (Kruckenberg and Haug 2006) will be adressed below, but first it is time to address Scandinavian sequence studies of the last century.

1.3.3. Scandianvian sequence research Most of the Scandinavian sequence research has focused on the sequences and repertories of Norway/Iceland and those of Sweden. It adds to the interest that Norway/Iceland and Sweden show large differences in their repertories and remaining sequences. “Scandinavia” is certainly not a unified area as far as the sequences are concerned.

47

Part I: To study manuscript fragments

The earliest and most extensive work is Carl-Allan Moberg’s Über die Schwedischen Sequenzen (1927). This study is impressive both for its method and for its early date. As this study was carried out during the early stages of the work on fragments, Moberg based his findings primarily on printed missals and graduals from the different Swedish dioceses, and a few late manuscripts. For this he was criticized by Toni Schmid, who asserted that the fragments would give a wider knowledge of the sequences in Sweden (cf. Björkvall 2006, 47). In Norway the focus of sequence research has traditionally been on two things: 1) When and by whom was St. Olav’s sequence written and what was it influenced by? 2) Which sequences were used in Norway? The discovery of the sequence Lux illuxit letabunda for St. Olav, patron saint of Nidaros, and of Lux illuxit... lux est nobis of St. Hallvard, patron saint of Oslo, may be said to be the starting-point of Norwegian sequence research (cf. Reiss 1912). Reiss suggested that Lux illuxit was composed in the last part of the twelfth century, by archbishop Eirik (1189-1205) (Reiss 1912, 17). Eirik Vandvik pointed out that all four bishops and archbishops who were linked to St. Victor, Paris, could have written the sequence (Vandvik 1941). However, although striking similarities can be found between St. Victor-sequences and Lux illuxit, it is stylistically not a sequence in the style of St. Victor, but has the characteristics of a sequentia nova (“transitional sequence”), with an inconsistency in the use of rhytm and rhyme and the size of the strophes. A project of a different type was undertaken by Erik Eggen. He was able to identify c. 150 sequences in fragments from 30 manuscripts housed in Norwegian collections and 18 in Icelandic collections (Eggen 1968). Eggen’s method was governed by his purpose, which was to identify the sequences in Nidaros and present them in such a way that they could easily be sung. Although he did not analyse the manuscripts, the edition of his work included plates of all the fragments he used, which means that it

48

1.3. The sequence: rise, fall and resurrection is still enormously valuable. Eggen could not know which sequences were prescribed in the official Nidaros ordinal and which were not, as the Nidaros ordinal was edited more than ten years after his death, and a large number of the sequences discovered by him in Norwegian and Icelandic manuscripts were in fact not prescribed in the Nidaros ordinal. The repertory of sequences in the edition of the Nidaros ordinal by Lilli Gjerløw (1968) was based on the sequence incipit in the few remaining witnesses to the Nidaros ordinal. She found that the ordinal prescribed 111 sequence titles, plus 7 marginal additions, in all 118 sequences. Not all sequences in the ordinal are found in the fragments, and a few are untraced, such as the sequences Ave presul and Laus Iohannis for St. Jón of Hólar and Iocundemur for the translation of St. Magnus of the Orkneys (Gjerløw 1968, 432).43 In 1988 Gjerløw wrote a paper which was never published, called Sequences 1988. Status quo. RA (available only at the NRA). Here she presented her new findings regarding Eggen’s sequences, and her own additional sequence titles found among Norwegian sources unknown to Eggen. The comparison of sequence titles in the fragments to those listed in the Nidaros ordinal is a good point of departure for studying the authority of the ordinal.

1.3.4. A new book on the Nidaros sequences The most recent chapter in the research of the Nidaros sequence repertory is the book The Sequences of Nidaros. A Nordic Repertory & Its European Context (Kruckenberg and Haug 2006). Several important aspects of the Nidaros sequences are addressed in this book, such as a comparison of the Nidaros sequence repertory with that of the rest of Europe regarding both size and contents, and contrasting the Nidaros sequences to those used in Swedish dioceses. 43

Kruckenberg lists six of the sequences which are untraced, and discusses whether two of them may be untraced due to corrupt entries in the ordinal manuscripts (Kruckenberg 2006, 18-19). Not included in this, the latest list of untraced sequences, is Precluis ecclesia, which should also be counted as untraced because of the hypothetical nature of the suggestion that it might be one of the unknown sequences in Seqv 25 (cf. Gjerløw 1968, 433).

49

Part I: To study manuscript fragments

An interesting addition to Lilli Gjerløw’s edition of the Nidaros ordinal is Lori Kruckeberg’s reconstructed sequence repertory for those parts of the ordinal which were not fully transmitted. There are two months for which the edition of the Nidaros ordinal had to rely on an abridged version, namely from the Nativity of the Virgin (8 Sept) to All Saints (31 Oct) (cf. Gjerløw 1968, 65-6). Kruckenberg suggests a reconstruction of the sequence repertory for these two months, based on the pattern of the ordinal for the remaining parts of the year. In addition to the four sequences present in the abridged manuscript for this period, it is likely that the original ordinal (i.e. without marginal additions) contained between six and twelve more sequences, making the total in the original ordinal not 111, but between 117 and 123 sequences. And with marginal additions of later sequences and feasts, the Nidaros ordinal did not contain 118, but 135 sequences (Kruckenberg 2006, 7-16, 34-5). In chapters by Kruckenberg, Hiley and Bower the Nidaros sequence repertory is compared to repertories in England and Germany in particular, but also France, in an investigation of the different layers of sequences in the ordinal and their relationship to different European regions. About one third of the sequences of the Nidaros ordinal may be labelled “German” (Bower 2006, 122). Of these sequences, some later spread to France and England, and may have been introduced to Norway from the South-West. However, some German sequences are otherwise transmitted in German sources only, and the path for these sequences seems to have gone directly from Germany to Scandiavia, and be connected with the Hirsau reform (cf. Kruckenberg 2006, 31-2). For example, a sequence like Grates honos is found only in manuscripts with an origin in southern Germany, most of which have a Hirsau affiliation,44 and in Nidaros.45 David Hiley has presented a survey of sequences in English sources. A large part of the sequences used in Nidaros belong to the general “Anglo-French” repertory. A few are very closely connected to England, as they are found only in

44

I thank Lori Kruckenberg for this information. Grates honos is represented in Nidaros through its presence in the ordinal, in an Icelandic fragment in Stockholm, and in a Norwegian missal (Seqv 18/Mi 106). The presence of Grates honos in this missal is first identified in a later chapter in this thesis.

45

50

1.3. The sequence: rise, fall and resurrection England and Nidaros (cf. Kruckenberg 2006, 29-30). In other words, the selection of sequences in the Nidaros repertory seems remarkable in its variety. Still, the organisation of the Nidaros ordinal does not seem random. Lilli Gjerløw was the first to remark on a peculiar feature of the ordinal, which consisted in the Notker-sequences being pushed back to “second place”, more precisely to the celebration of the octave of a feast, while Anglo-French sequences were used for the main feasts (Gjerløw 1968, 434). Kruckenberg analyses this feature thoroughly, and finds that it is not just the case for the saint’s feasts mentioned by Gjerløw, but for thirteen important feasts in all, like Epiphany, Easter, the Ascension and Trinity Sunday (Kruckenberg 2006, 32-3). She assigns this to the conscious deliberations of the ordinal’s writer or writers, who reduced the German practice in favour of a repertory of sequences arriving from England (ibid., 27-8). The status of Swedish sequence research is described by Gunilla Björkvall in the chapter “Sequences in the Fragments at the Swedish National Archives” (pp. 45-62). The differences between Norway/Iceland and Sweden seem more striking than the similarities. In Sweden the seven dioceses do follow one liturgical rite, but have their own liturgical traditions. This means that there is not one repertory of sequences, but several, connected to the single dioceses and religious orders (Björkvall 2006). While Norway has fragments from approximately 70 manuscripts with sequences, Sweden has fragments from 448 manuscripts, of which 243 are sequentiaries 108 missals and 78 graduals (ibid, 47). While the majority of these are late (1400 or later), 115 are from the fourteenth century, 61 from the thirteenth, and 19 are from the twelfth. Fragments from almost twice as many twelfth-century manuscripts are preserved in Sweden than in Norway, and for the following centuries the difference increases. In all 250 sequence texts are identified among the Swedish fragments, of which 77 are not found in Moberg. This is seen to confirm Toni Schmid’s suggestion that the fragments would contribute a great deal to sequence research, and form a broader picture than when merely using late and printed sources. Of the sequences present in Swedish sources, 142 do not appear in Norway. Similarly, of the sequences collected

51

Part I: To study manuscript fragments by Eggen and Gjerløw 85 titles do not appear in Sweden. These are significant numbers. The clear difference in the selection of sequences in the eastern and western parts of Scandinavia is underlined by the fact that the Swedish ordinals and fragments generally do not prescribe sequences for Advent (in accordance with the German tradition), while the Nidaros ordinal and the Norwegian manuscripts generally do (in accordance with the Anglo-French tradition). Yet another contrast between Norway and Sweden is that while the sequences in 46

Norway believed to be local products almost can be counted on one hand, Moberg assumes that c. 60 sequences have a Swedish origin (although none before 1300) (cf. Björkvall 2006, 56). A considerable number of sequences are also connected to the Finnish diocese of Åbo (ibid.). The studies in this book do not only investigate influences on Nidaros through repertorial layers and the mere presence of sequences, but also through in-depth studies of single sequences. Rebecca Maloy has taken a closer look at Sancti baptiste as it is transmitted in NRA, Lat. fragm. 418 (Seqv 1), NRA Lat. fragm. 627 (Seqv 5) and a fragment of Icelandic origin (Copenhagen, AM 241 A III Acc. 7b). She has found that these remaining Nidaros sources present a mixture of Norman verbal traits and German melodic characteristics, and suggests that the reason for this “particular blend” can be found in Nidaros’s historical associations and patterns of cultural influences. The German melodic features support the hypothesis of an East-Frankish repertory entering Nidaros in the late eleventh century when it was a province of Hamburg-Bremen (Maloy 2006, 260). Other studies of sequences also present in NRA Lat. fragm. 418, like Clare sanctorum (Snyder and Altstatt 2006) and Sacrosancta hodierna (Zimmermann 2006), clearly demonstrate that when the sources are in a fragmentary state, it is often not possible to assign the fragments to any tradition based on variants in text or melody (see, for instance, Snyder and Altstatt 2006, 225-6). 46

Lux illuxit letabunda (St. Olav), Predicasti dei care (St. Olav), Postquam calix babylonis (St. Olav), Lux illuxit... lux est nobis (St. Hallvard), and possibly Laudes debitas deo (de una virgine) and Quare fremuerunt gentes (unknown, but suggested used for the Holy blood, cf. Kruckenberg 2006, 35).

52

1.3. The sequence: rise, fall and resurrection

Kruckenberg has studied the “sequentiae novae” Celeste organum and Stola iocunditatis and found that they went north following very different paths, based on a comparison of the versions found in Nidaros and a broad spectrum of sources from the rest of Europe. While Celeste seems to have come through England, Stola iocunditatis as presented in a thirteenth-century breviary-missal (Seqv 38 add/Br-Mi 3) was probably not transmitted through English or French sources, but from the eastern parts of Europe, with possible connections to the Premonstratensians. It also seems clear from the Swedish evidence that Stola iocunditatis did not reach Nidaros by going through Sweden (Kruckenberg 2006b, 401). Bower, who studied the German sequences of Nidaros, attempted the outline of a historical scenario: in the eleventh and twelfth centuries a sequence repertory from Germany formed the foundation of the genre, while a new repertory of English and French sequences was introduced to an independent arch see in the twelfth and thirteenth century (Bower 2006a, 128). Although Bower emphasised the preliminary nature of this suggested course of events, the oldest evidence of liturgical manuscripts speaks against a theory of “first one, then the other”. Although the earliest surviving missal fragments do not include sequences, the main characteristic of the eleventh century manuscript material is a peculiar mixture of English and German influences (cf. Karlsen 2003). As the fragments from the earliest missals show English and German impulses in Norway in the late eleventh century, both leaving their imprints on the same manuscript pages, there is no reason why this should not also be the case for the sequences. And if the English and/or French and German sequences had existed side by side for almost a century by the time of the establishment of the archbishopric, the task of putting together an official sequence repertory would be all the more challenging. The Nidaros ordinal may be the work of a man who used and preferred Anglo-French sequences, but he may not have been able to disregard the demands of a clergy used to having a wide variety of sequences, some of them fonder of the German tradition. If so, the Nidaros ordinal would be an attempt to create “order”, but still keep in use the sequences people were most attached to. Another possibility is that the ordinal was the answer to a gradual shift

53

Part I: To study manuscript fragments in the power-balance, and that Anglo-French influence increased during the course of the twelfth century, gradually pushing the German sequences back and therefore giving the Notker-sequences the secondary role they play in the ordinal. Such a scenario might find support in one of the pre-ordinal sequentiaries in this study containing both English and German sequences for Christmas, since the Notkersequence already here plays a secondary role (Seqv 40). As the discussion above demonstrates, the manuscripts themselves, even in the form of fragments, can contribute to the interpretation of musicological evidence. The Sequences of Nidaros demonstrates that musicologists can identify layers of influence from different European regions in a sequence repertory as well as in the textual and melodic variants of transmitted sequences. The aim of the current study is to recognise similar layers of influence in the palaeographical evidence and identify connections to different European traditions as they are traceable in the manuscripts.

54

1.4. The archbishopric of Nidaros

1.4. The archbishopric of Nidaros The remains of the books, the scribes writing them and the contents of these books are all part of a historical context. It is therefore difficult to study them without an eye on historical circumstances. The most relevant research for this study is that on Nidaros, the bishops’ sees and the monasteries.

1.4.1. Nidaros The town of Nidaros (current Trondheim) emerged as an important centre in the North not long after the violent death of king Olav Haraldsson (d. 1030) in the battle of Stiklestad. Olav’s potential for sainthood seems to have been immediately recognised by one of his own bishops, the Englishman Grimkell (d. 1047),47 whose involvement in the translation of Olav is described in medieval sources such as Snorri Sturlusson’s Saint Olav’s saga (Aðalbjarnarson 2002 (1945), 403-4) and The Legendary Olav’s saga (Johnsen 1922, 90). The first liturgy for the feast of St. Olav 29 July was put together within decades of his death on the basis of the common material for saints.48 The identity and growth of Nidaros as a religious centre in the North was highly dependent on the creation of Scandinavia’s first known local saint (for different aspects of local and universal sainthood, see Thacker and Sharpe 2002). Olav was a royal warrior and martyr in the tradition of English saints like Oswald (d. 642) and Edmund (d. 869).49 His effigy is an imitiatio of the victorious Christ in his Romanesque form, the triumphant heavenly king. The cult of St. Olav did not only spread to the rest of Scandinavia, but also to England, probably due to Grimkell (Østrem 2001, 30).

47

For Grimkell, see Johnsen 1975, Østrem 2001, 28-33. For other English bishops in Norway at the time of Olav Haraldsson, see Lapidge (2003, 55). 48 The oldest evidence for the proper mass of St. Olav is the so-called Red Book of Darley (Cambridge, CCC 422) from the early 1060’s. The first proper office is in the manuscript known as the Leofric Collectar dated c. 1050-60 (London, British Library, Harley 2961), cf. Gjerløw 1968, 124-8. For the oldest office of St. Olav, see Østrem 2001, 28-31. 49 These two English royal saints were also remembered in the later Nidaros ordinal, Oswald with a commemoration on 5 August and Edmund with his proper office 20 November. Collects from their offices were also used in the earliest office of St. Olav (Gjerløw 1968, 125).

55

Part I: To study manuscript fragments The growth of Nidaros as a Christian centre led to its change of status from a provincial bishops’ see to arch see in 1152/53. The new archbishopric comprised the Norwegian mainland (including Bohuslän in present day Sweden) and the western isles (Iceland, Greenland, the Faroe Islands, Shetland, the Orkneys, Sudor and Man). Before becoming an independent archbishopric Nidaros had been a province of Hamburg-Bremen (c. 1043-1103) and of Lund in Denmark (1103-1152/53). In spite of the formal connection to the south, the process of Christianisation, starting in the tenth and becoming established during the eleventh century, was a complex one. While some missionaries and bishops were sent north from Hamburg/Bremen, others had followed the converted kings like Olav Tryggvason (d. 1000) or Olav Haraldsson from England. The various influences from different European regions are embedded in the medieval liturgy. Both the evidence of the pre-ordinal use(s) of Nidaros (Gjerløw 1979) and the Nidaros ordinal (Gjerløw 1968) show a clear link to Lund, but also to uses in England, France and Germany. Gjerløw noticed the similarities between the liturgy of Lund, the ordinal of Nidaros and an ordinal from the reformed Hirsau abbey in southern Germany, as well as connections to rites in England (Gjerløw 1968, 85-110). New initiatives in architeture, literature, liturgy and historiography seem to have been spurred on by Nidaros’ new status in the second half of the twelfth century, many of them around the time of the prominent figure of Archbishop Eystein Erlendsson (1161-88).50 The activity in the last half of this century included the building of a new cathedral in Nidaros, the new office In regali fastigio (ed. Østrem 2001) based on the Passio Olavi (ed. Metcalfe 1881), and the sequence Lux illuxit for St. Olav (ed. Reiss 1912), as well as historical works like Theodericus Monachus’ Historia de antiquitate regum Norwagensium (ed. Storm 1880) and Historia Norwegie (Ekrem and Boje Mortensen 2003).51 This was a time of hectic activity for scribes and book50

Eystein is the subject of several studies (Vandvik 1961; Gunnes 1996). For Eystein’s connection to St. Victor, Paris, see Johnsen 1943-46. 51 The role of Nidaros in the literature of St. Olav has been treated in a recent article (Mortensen and Mundal 2003). The initiatives in Nidaros were followed by the writing of Latin legends for the saints of Selja (8 July) in Bergen and St. Hallvard (15 May) in Oslo (for the four versions of St. Hallvard’s legend and a possible office, see Gjerløw 1968, 421-424). The importance of Latin as a literary language has at times been underestimated (cf. Svensson 1974, 200), a view challenged by Lars Boje Mortensen in for instance “Den norske middelalderlitteratur på latin”, 139-146 (Imsen 2005). The very haphazard

56

1.4. The archbishopric of Nidaros binders, whether they were copying the latest literature or liturgical books for use in the many churches and chapels. The sequences and the production of liturgical manuscripts were an integral part of the increased awareness and self confidence of the new archbishopric. It is likely that most of the scribal activity both before and after the independence of Nidaros was connected to the bishops’ sees and monasteries, which is the topic of the next chapter.

1.4.2. Bishops’ sees, monasteries and scriptoria The locally produced manuscripts in the centre of this study are probably made in connection with religious institutions, such as the bishops’ sees and chapters, monasteries or convents, mainly because of their liturgical contents. It is therefore worth taking a closer look at the specific centres where books were kept and scribes could work. Norway had a relatively small population and few towns. The important centres were: 1) The arch see and the bishops’ sees with cathedrals, chapters52 and priests: Nidaros, Bergen, Oslo, Stavanger and Hamar. 2) The religious houses, mainly Benedictine, Cistercian, Augustinian, Dominican and Franciscan, either along the coast or in proximity to the arch see or bishops’ sees, and in the towns Tunsberg and Konghelle (see map below).53 3) The royal court (pre 1300 mainly situated in Bergen, later in Oslo) and the aristocracy.

transmission of several Latin texts, often with only one manuscript witness (or indeed only later printed evidence), makes it likely that a rather large portion of the original Latin literature is lost. Still, the texts which are transmitted testify to a conscious rendering of the past and local myth-making, as shown in a recent comparative study treating the first literature in Norway, Denmark and Hungary (Mortensen 2006). 52 It is assumed that the cathedral chapter of Nidaros had c. 24 canons, while Bergen, Oslo and Stavanger had c. 12 canons. The smaller see of Hamar seems to have had a slightly smaller number (Hübert 1922). 53 The sources of monastic life in Norway are the archaeological sites and more than 500 documents about the monasteries still remaining, either in the form of the medieval originals or through later copies. The documents are published both in book form (Diplomatarium Norvegicum (DN) 1849-) and on line: http://www.dokpro.uio.no/dipl_norv/diplom_felt.html. None of the sites or documents reveal anything about the thing that interests us the most in this context, namely book production.

57

Part I: To study manuscript fragments

The origin of writing in Norway can probably be found in connection with the first bishops’ sees and monastic foundations. Several of the earliest bishops were foreigners, and most of the religious houses in medieval Norway were also founded by the help and participation of foreign houses. These were probably among the first men who provided the know-how, means and circumstances for local bookproduction of a stable and efficient nature. The role of the first Benedictine monasteries established in connection to bishops’ sees has not been thoroughly investigated. Knut Helle suggests that although the level of learning in the first monasteries may have been modest, they still functioned as bridgeheads for the knowledge of reading and writing. He emphasises the important role of the monasteries and other religious houses, like the Augustinians or regular canons, in the production of new literature in Latin and Old Norse (Helle 1993, 114). In a recent study on the Norwegian medieval literature in Latin by Lars Boje Mortensen the bishops and regular canons are referred to as the primary distributors of foreign learning and liturgy to Norway. According to Mortensen literary activity was connected to the towns, to court and bishop, more than monasteries (Mortensen 2005, 145). Although the character of scholarly activity and book production in a Benedictine or Cistercian community may have differed from the activity in the bishop’s quarters, the chapter or an Augustinian house, they have all been a part of the dynamics of the town and the bishop’s see. As most Benedictine or Cistercian monasteries were either royal/aristocratic or episcopal foundations, they had natural ties to bishop and court. It is important to consider all the relevant institutions, bishops’ sees and different religious houses alike, as they may have played different but vital parts in the production, preservation and distribution of books, literature, liturgy and learning. In the field of archaeology Alf Tore Hommedal emphasises the need to look at the centres of scribal culture as connected entities, and not as separate institutions (Hommedal 2006, 85). Although too little is known about the dynamics in the towns, between bishop, secular canons of the chapters, regular canons of Augustinian houses, Benedictines and other religious orders, these dynamics should be considered in the study of medieval scribal culture.

58

1.4. The archbishopric of Nidaros The conditions under which the manuscripts were made would have probably been quite different. A medieval “scriptorium” may have had different realisations, and not necessarily within the walls of a monastery. In Norway very little is known about how many of the medieval books were actually produced inside monastery walls and how many were made in connection with bishop or chapter, or how the commission of a book would work in practice. Although there is no evidence of secular workshops for book production in Norway before the Reformation, we cannot rule out the existence of lay-men involved in the business of book production from c. 1200 onwards. Scribes would in all likelihood be available outside the monasteries, among the men of the bishop, in the royal court and possibly as freelancers in secular workshops during most of the Middle Ages. Not all scribal activity would necessarily require stable environments, for instance the drawing up of a document or writing of a letter. The copying of a missal or another book, however, would require proper working conditions. More is known about the working conditions inside a monastery than outside, although scribal practice has played a very peripheral role in Norwegian research on monasteries. The basic monograph on the Norwegian monasteries is still Chr. C. A. Lange’s De norske Klostres Historie i Middelalderen (1847). A yearbook from “Foreningen til norske fortidsminnemerkers bevaring” for the year 1987 is a modern attempt to give a collective survey of all the monasteries in Norway, and it includes additional information gathered through more than a century of archaeological excavations and scholarly research. Other important studies for monasticism in the North are Tore Nyberg’s Monasticism in North-Western Europe, 800-1200 (Nyberg 2000) and James France’s The Cistercians in Scandinavia (France 1992). Monasticism in Norway seems to have been of an international character throughout the Middle Ages. While the Benedictine and Cistercian monasteries from the first half of the twelfth century were primarily founded from England, the Augustinian houses, which archbishop Eystein seems to have promoted, were probably connected to his past as a canon in the Augustinian house of St. Victor in Paris (Gunnes 1996).

59

Part I: To study manuscript fragments The houses founded with the help of English monks kept in close contact with England well into the thirteenth century and possibly after 1300 (Gunnes 1987, 55). The Norwegian houses varied in size, and the reported numbers of resident monks and nuns extend from the single regular canon reported to be left in St. John’s in Bergen in 1436,54 to the 85 sisters and brothers belonging to the damaged Brigittine monastery of Munkeliv in 1462.55 Between ten and twenty monks or nuns seem to have been a “normal” size Norwegian religious house, at least for the twelfthfifteenth century.

56

A monastic scribal community could be small or large, well equipped or poor, depending on the size and financial situation of the monastery. A scriptorium - in the sense of the scribes’ workspace - had relatively few physical requirements. A room could without much effort be turned into a scriptorium for a limited period of time. This lack of physical characteristics of a scriptorium makes it difficult to identify any in the remaining walls of the Norwegian monasteries. The furniture would be wooden and is long gone (a wooden chair and desk, with the top plate tilted towards the scribe, is the furniture displayed in medieval illustrations). It has been assumed that monastic scribes would find it convenient to write outdoors, in the cloister. What would be a perfectly good workplace in Italy or France may not be so convenient on the cold and windy coast of western Norway. A gush of wind catching a stack of parchment leaves or turning over an inkhorn could lead to the waste of precious materials, not to mention the possible ruin of days or weeks of hard labour. The mere weather conditions would suggest that Norwegian monasteries involved in book production would require indoor spaces where the scribes could work, protected 54

DN 17, 536 DN 17, 567. The transition of Munkeliv from a Benedictine monastery is well documented in the DN, see for instance DN 4, 809-14, and also the article on the matter by E. Gunnes (Gunnes 1990). The Cistercian period from 1455-78, when the nuns of Munkeliv changed places with the monks at Hovedøya, is testified in DN 17, 657, DN 21, 563 and DN 7, 481. 56 When the Cistercian monastery Lyse south of Bergen was founded from Fountains, England, abbot Ranulph was probably followed by twelve monks, as laid down in the Cistercian statutes, paragraph xii: “Duodecim monachi cum abbate tertio decimo ad cenobia nova transmittantur” (Guignard 1878) . Hovedøya, also Cistercian, is reported to have twelve monks in 1462, but with room for twenty (DN 17, 657). At one point all the Dominicans in St. Olav’s in Oslo are named, and then they were twelve (Hommedal 1986). 55

60

1.4. The archbishopric of Nidaros from wind and rain and with a fair amont of daylight and warmth. Alf Tore Hommedal has in a recent article about Norwegian scriptoria pointed out that to search for one fixed room called “scriptorium” may be futile. Rooms in a Norwegian monastery could have had multiple functions, and it is not unlikely that scribes could work in the library during summer and in the winter move to the calefactorium (Hommedal 2006, 87). That a scriptorium further south could be vulnerable to the hardships of winter is evident in a letter from Cuthbert, abbot at Wearmouth-Jarrow in 763-4: The conditions of the past winter oppressed the island of our race very horribly with cold and ice and long and widespread storms of wind and rain, so that the hand of the scribe was hindered from producing a great number of books. (quoted from Parkes 1982, 15)

As Parkes has demonstrated in the case of Wearmouth-Jarrow (Parkes 1982), initiatives to book production may have come both from needs or desires within the house or from external requests. Evidence from other parts of Europe reveals that a monastic scriptorium would not be isolated from people outside the community or related institutions. A medieval book could be the result of co-operation – between monks and lay-men, between monks of different orders, and even between monks and nuns, as documented in the twelfth century (Alexander 1992, 16-20). One may say that the history of books and the attempts to reconstruct medieval book culture have developed into different fields according to emphasis: Some deal primarily with the literature or liturgy, others with the book production and some with book collections. Ludvig Holm-Olsen wrote what may be said to be the summary of what was known about Norwegian book collections and book production in the last century, in his Med fjærpenn og pergament (1990). The traditional sources now appear to be “drained”, and references to the same documents and inventories are repeated (as they will be once more below). The fact that Norway is left with the “short end of the stick” compared to other Nordic countries with respect to remaining medieval manuscripts, is a general cause for lament in book and library

61

Part I: To study manuscript fragments studies, and rightly so one might say. Still, as the remains of more than a thousand books are kept in the NRA, these do represent evidence of both book production and book collections, and can probably to a large degree provide new information about the scribal activity and learning within monastic walls as well as in other institutions connected to court and bishop. To get further it is necessary to study the remaining manuscript material in detail, and try to relate it to the historical context. Below is a map of monasteries and the mainland bishoprics, and a survey of Norwegian monasteries.

62

1.4. The archbishopric of Nidaros

. Map 2: Main land bishoprics and male and female religious houses in medieval Norway. Northern Norway had no known monasteries, and is therefore not included in the map. The larger stipled line marks the medieval border, and the shorter stipled line is the present border.57 57

I thank Alf Tore Hommedal for providing electronic images of maps from two of his articles (Hommedal 1999a, 1999b), which are here combined into one.

63

Part I: To study manuscript fragments 58

Table 2: Religious houses in medieval Norway I. Benedictine houses • Nidarholm59 (St. Laurentius) • Selja (St. Alban’s) • Munkeliv (St. Michael’s) • Gimsø* • Nonneseter* (St. Mary’s • Bakke* • Nonneseter* (St. Mary’s)

Place Island by Nidaros Island by Stad Bergen Island by Skien Oslo

After 1111-c. 1540 Bef. 1161-154761

Nidaros, east of the river Bergen

Bef. 1150-1537 (robbed 1563?) A. c. 1150-1507 B. 1507-1528

II. Cistercian houses • Lyse • Hovedøy (St. Edmund’s and St. Mary’s) • Munkeby • Tautra

Place Os, south of Bergen

Date 1146-1536

Island by Oslo

1147-1532

Levanger The island Tautra

Bef. 1180 1204-1531

III. Augustinian houses • St. Olav’s • Halsnøy (Holy Spirit) • Helgeseter • Kastelle • Rein*64 • St. John’s65

Date A. 1028? B. c. 1100-1537 1100-between 1461 and 74 A. 1110-1425 B. 1425-1531

Place Stavanger

Island between Stavanger and Bergen Nidaros Konghelle (now Sweden) Rissa Bergen

Date Bef. 1160after 1236 1163/64-1536 Bef. 1183-1546 Bef. 1181-1529 After 1226-1532 Bef. 1208-1425

58

Founder A. King Knut the great60 B. Sigurd Ullstreng (Nobleman) King Eystein (r. 1103-1123) Brigittine 1425-55 + 1478-1531, Cistercian 1455-78) Nobleman Dag Eilivson

A. Formerly assumed to be Cistercian.62 B. Taken over by Antonites

Founder Bishop Sigurd of Bergen, with monks from Fountains, England. Bishop Vilhjalm of Oslo (?), with monks from Kirkstead, England. founded by monks from Lyse, Bergen.

Founder Assignment uncertain. Possibly Benedictine. Erling Skakke (d. 1179), with help from Arcbishop Eystein (1161-88) and Wellow by Grimsby?63 Archbishop Eystein Archbishop Eystein Skule jarl Bårdsson (1189-1240)

This survey is made using Foreningen til norske fortidsminnesmerkers bevaring, yearbook 1987, although the information on Nonneseter in Bergen is supplemented using Diplomatarium Norvegicum (on line). 59 Nidarholm seems to have had an association to the Cluniacs of uncertain formality (Lunde 1987). 60 Matthew Paris is the only source for this information (Gunnes 1987, 51). 61 Even though the monastery became a fief in 1547, there were still nuns as late as 1586 (Gunnes 1987, 81). 62 The arguments in support of Nonneseter being a Cistercian nunnery are so questionable and the counter-arguments so convincing (Tryti 1987), that this alone would suggest that Nonneseter was a Benedictine house. That Nonneseter in DN 17, 799 is referred to as “ordinis sancti Benedicti” should be considered the decisive argument for Nonneseter being Benedictine, and not Cistercian. 63 The abbot from Wellow by Grimsby attended the crowning of Erling’s son Magnus in Bergen in 1163/64 (Gunnes 1987, 58). 64 Of uncertain order.

64

1.4. The archbishopric of Nidaros

• Utstein (St. Laurentius’)

Place The island Mosterøy, north of Stavanger

IV. Premonstratensian houses • St. Olav’s • Dragsmark, (”Mariskog”)

Place Tønsberg Bohuslän (now Sweden)

V. Dominican houses • The Dominican convent • St. Olav’s • The Dominican convent • St. Olav’s66

VI. Franciscan houses • The Franciscan convent • St. Olav’s • The Franciscan convent • The Franciscan convent • The Franciscan convent • The Franciscan convent

Date After 1263-1536

Founder King Magnus Lagabøter (r. 1263-80)

Date Bef. 1190-1532 1236? (At least bef. 1260)-1532

Founder King Håkon Håkonson (r. 1217-63)

Place Nidaros Oslo Bergen Hamar

Date Bef. 1234-1537 1239-1537 Bef. 1247-1528 Mentioned 1511

Place Tønsberg Bergen Konghelle (now Sweden) Oslo Marstrand (now Sweden) Nidaros

Date Bef. 1236-1536 1240’s-1536 Bef. 1272-1532 Bef. 1291-1537 Bef. 1291-1532 Bef. 1472-1537

VII. Other houses Iohanittes: • Varna

Place South of Moss

Date Between 1170 and 1270-1532

* = female communities

65

That St. John’s should be a house of “barefoot-brothers”, or so-called “discalced Augustinians”, as is sometimes claimed, is not supported in the sources, and is even less likely since this is a later reformmovement. It is more likely that there has been some confusion with the Franciscans, which were regularly called “barefoot-brothers” in the sources (for instance in DN 15, 1, where they are mentioned in a will right after St. John’s church: “st: Johannes kierche j # brende, Barföder bröder i Bergenn x # brende”). That they were regular canons is attested to in DN 17, 535 (monasterium sancti Iohannis baptiste ordinis sancti Augustini canonicorum regularium). See Nenseter 2003, 53. 66 The monastery may have been used by the Dominicans, and it may also have been used by Antonites for some time (Gunnes 1987, 59).

65

Part I: To study manuscript fragments

1.4.3. The date of the Nidaros ordinal The final chapter in this section will briefly address the question of the date of the Nidaros ordinal, as different ones have been proposed in the literature on the ordinal. Several of the manuscripts dealt with in this study have their origin in the first half of the thirteenth century, around the time of the introduction of the ordinal. Therefore the date of the ordinal is of crucial importance in the evaluation of the manuscripts. The Nidaros ordinal is one of the initiatives connected to Archbishop Eystein. During the early 1170s he was working on a unified liturgy for the Nidaros archbishopric, an ordinal (Gjerløw 1968, 87-8). However, he did not live to see the completion of the work, as he is referred to in the ordinal manuscripts as bone memorie (ibid., 322) and venerande memorie augustinus archiepiscopus (ibid, 387). The last authority to be mentioned, Archbishop Eirik (1189-1205, d. 1213) is called dominus archiepiscopus, which could indicate that he was still alive when the ordinal was completed (Gjerløw 1968, 30 and 386). Gjerløw sets the terminus post quem of the Nidaros ordinal to 1205, evidently assuming that the reference was made by Eirik' s successor. This might be the case, but the reference to dominus eiricius archiepiscopus could also have been made by a person entrusted with continuing the work on the ordinal because of Eirik' s exile in Denmark and later blindness in the 1190’s. The ordinal refers to a synod (ibid. 386) which was apparently organized before Eirik’s exile in Denmark 1190-1202, and even though it may be likely that the work was not finished until Eirik’s return from exile, the Ordo' s terminus post quem must, strictly speaking, be put to after 1190. The terminus ante quem was by Gjerløw in 1968 set to 1224, the date of the episcopal statutes of Bishop Magnús of Skálholt (1216-37). She later pointed to a secundum ordinem manuscript which could hardly have been written after 1220 (Gjerløw 1979, 228). Therefore the Nidaros ordinal appears to have a terminus post quem of 1190, and

66

1.4. The archbishopric of Nidaros a terminus ante quem of 1220, with the likelihood of a date some time between 1202, archbishop Eiric’s return from exile, and 1213, the year of his death.

67

These past chapters have dealt with more than a century of research in several fields, all in some way connected to fragment study and to this study in particular. I hope to make a further contribution in fragment studies, palaeography, sequence studies and the knowledge about the book and scribal culture in medieval Norway. Before moving on to the analysis, two other important aspects of fragment studies should be adressed. First I will make an assessment of how large a part of the medieval book culture we can ever hope to get a glimpse of, and how much is irretrievably lost. Secondly, I will discuss different approaches to fragment studies, and present the method used in this study, which can hopefully also be applicable for future research.

67

In her unpublished work The Sequences Status quo 1988 – RA Lilli Gjerløw has narrowed the date of the Nidaros ordinal to c. 1205-1213 (Gjerløw 1988, 1).

67

Part I: To study manuscript fragments

2. The remains of a book culture Although there are some codices remaining from medieval Norway, both imported and locally produced, the reality is that we, to an even larger degree than our Nordic neighbours, rely on fragments to provide glimpses from the book culture which once was introduced to these parts and adapted here. For sequences the case is that not one single handwritten sequentiary, gradual or missal remains from the Norwegian Middle Ages, either imported or locally produced. The earliest evidence of a complete missal with sequences is the printed Missale Nidrosiense (1519). How does the situation today relate to the medieval state of things? In this section the following questions will be asked and attempted answered: -

How much of the medieval corpus of books remains today through codices or fragmentary evidence? Which consequences should the transmission have on the conclusions we draw?

-

To what extent are liturgical manuscripts representative for medieval book culture?

-

To what extent are the fragments we have today from liturgical books representative for the liturgical manuscripts which were once in Norway in the Middle Ages?

2.1. Remains and losses - an estimate The purpose of this chapter is to discuss how much of the manuscript material is left compared to what existed at one point in time, or rather two selected points: c. 1300 and the time of the Reformation (1537). While the liturgical books lost their value more or less at the snap of a finger, there also seems have been a slower process starting with the introduction of the printed book in the fifteenth century, changing how people regarded handwritten books. The Norwegian book corpus in the Middle Ages was a complex one, with books for different purposes, some in Latin, some in Old Norse. In order to make a general estimate, the distinction between library books and liturgical books used by for instance Michael Lapidge can be clarifying also for our purpose. The “library books” 68

2.1. Remains and losses - an estimate are those placed and used in a library, by Lapidge defined as “a collection of books acquired and arranged for the purposes of study and the pursuit of knowledge,” a definition which excludes the liturgical books (Lapidge 2006, 1). The “liturgical books” are books of various genres needed for the celebration of mass and office, mainly connected to the churches and to a larger or lesser degree to private worship. One type of book, which strictly speaking does not come within Lapidge’s definition, seems to have been in existence for as long as people have been reading, namely the book acquired and read for entertainment, like the romance novels or “sagas”. Since it would be difficult to draw a line between some of these and some of the library books, I include the novels and sagas in the library book “rubric”. While the liturgical books were all Latin, library books and fine literature could be in either Latin or Old Norse. In other words, we can divide the Norwegian book corpus into liturgical books and library books, and the library books into those written in Latin, and those in Old Norse. These groups would include both imported and locally produced books. Latin books could come from all over Europe, while a considerable amount of the Old Norse books were imported from Iceland (cf. Karlsson 1979).

2.1.1 What is left? To start with the largest collection at the NRA, one may say in a rough estimate that of the c. 1200 medieval manuscripts found among the fragments in the NRA, about one thousand are liturgical books, approximately one hundred are Latin books of different genres and probably less than a hundred are Old Norse books.68

68

From the beginning there seems to have been a basic organization of the Latin fragments according to contents, starting with different types of prose texts. The c. 60 first envelopes contain authors like Isidore of Seville (Lat. fragm. 1-3) Petrus Comestor (Lat. fragm. 8 and 16), Leo the great (Lat. fragm. 26), Petrus Lombardus (Lat. fragm. 47), Augustine’s De civitate dei (Lat. fragm. 51), Gregory’s Moralia in Job (Lat. fragm. 60), just to mention a few identified examples. Some Latin prose texts are also registered among the Old Norse fragments, like a fragment from Sallust’s Iugurtha (Old Norse fragm. 93). Gjerløw registered works of this type under the heading “Auctores” (see for instance a brief survey of the non-liturgical works in Karlsen 2003, 61). From Lat. fragm. 96 onwards the fragments in the NRA appear to be mainly from liturgical manuscripts.

69

Part I: To study manuscript fragments The few Latin codices which remain are in several collections, mainly in Sweden and Denmark. Very few books remain from before 1200. One exception is an English twelfth-century Gospel-book from the Benedictine monastery Munkeliv in Bergen (Copenhagen, Royal Library, GKS 1347, 4°). On the last page is written a property-list which is one of the oldest remaining examples of Norwegian vernacular script (Hødnebø 1960, plate 1). The other book from just before or around 1200 is a rather unique psalter, called the Kvikne Psalter after the place of its secondary provenance (Oslo, National Library, Ms.8° 102). Although badly damaged, it is still in its medieval binding, and is the oldest remaining codex written in Latin. Remarkably enough it has completely escaped notice, with the exception of the runes on its medieval wooden cover (Olsen 1960, 162-3). Three other books are also psalters, but of a different character: Two thirtheenthcentury high quality psalters imported from England and France belonged to Norwegian noble women, and are now in Copenhagen and Berlin. A fifteenth century psalter (now in Prague) is evidence of the book production at at Munkeliv in Bergen, at that time Brigittine. One extraordinary medieval book is Aslak Bolt’s Bible, copied in France in the thirteenth century, now in Oslo. Three books presumed to be of Norwegian origin are in fact composite codices (Copenhagen, Royal Library, NKS 32 8°, NKS 133 4° and Thott 110 8°), edited in Manuale Norvegicum (Fæhn 1962). These nine remaining Latin books are all liturgical – a Gospelbook, psalters, manuals, a Bible – but apart from the manuals they are not what may be called “rite-specific” (i.e. the contents are not governed by an ordinal or particular rite, but are the same all over Europe) and, again with the exception of the manuals, they do not contain chants, which means that they have not been valued as interesting for liturgical or musicological studies. There are also parts of the Old Norse book corpus remaining, and many of the medieval manuscripts in the vernacular have survived in the form of codices. Anne Holtsmark wrote an informative survey of remaining Old Norse literature and the historical reasons for the losses in the article En side av norsk bokhistorie (Holtsmark

70

2.1. Remains and losses - an estimate 1956a). There is a list of medieval manuscripts containing Old Norse prose in the Register volume of the planned Dictionary of Old Norse, and in some cases provenance is suggested (Den Arnamagnæanske Kommission 1989). According to Odd Einar Haugen c. 125 books, the larger part of which are lawbooks, remain in the Norwegian vernacular, mainly as codices, but some also as fragments. He mentions that at least twenty books are of other genres, of which three will be included in the table below, as they were written by scribes who otherwise wrote books in Latin (cf. Gjerløw 1968, 34-8). These three books are the Old Norwegian Homily Book (Copenhagen, AM 619 4°), Konungsskuggsjá or The king’s mirror (Copenhagen, AM 243 b

fol.) and the Legendary Olav’s saga (Uppsala, DG 8 II). While the two first

books are connected with Bergen (Tveitane 1981, 101 and 104), the last was presumably written in Nidaros (Holm-Olsen 1990, 99). The three Old Norse books just mentioned and more than fifty other books (not including law books) are listed in an article by Stefán Karlsson on the Iceland book exports to Norway (Karlsson 1979). Karlsson lists thirty books of Icelandic origin known to have been in Norway in the Middle Ages, a testimony to the large number of Icelandic books kept in Norwegian collections at that time.69 If we add the c. 30 Icelandic books from Karlssons list to the 125 estimated Norwegian books, c. 155 of the Old Norse books in Norway in the Middle Ages remain either as codices or fragments, most of which are kept in Copenhagen. Most of the Old Norse books are not included in the table below, as it would make it too extensive. Only the books whose scribes also wrote Latin books are included.

69

Karlsson’s method for assigning books to Icelandic scribes met objections from Magnus Rindal (cf. Kyrkjebø 2003, 24-28).

71

Part I: To study manuscript fragments Table 3: Collections holding Latin books, or Old Norwegian books with scribes also writing in Latin, present in Norway in the Middle Ages: 12th c. Berlin Kupferstichkabinett Copenhagen, the Arnamagn. Collection

Copenhagen, Royal Library

Oslo, Deichman Library Oslo, National Library

• GKS 1347 4º (Gospelbook. Origin: England) • NKS 32 8° Manual. Latin. (Origin: Norway)

13th c. • Psalter of Margrete Skulesdatter (Origin: London) • AM 243b , fol. The king’s mirror. O. N. (Origin: Norway) • AM 619 4º (Homily book in Old Norse. Origin: Bergen) • GKS 1606 4º (Psalter. (Origin: Paris) • NKS 133 4° Manual. Latin. (Origin: Norway) • Thott 110 8° Manual. Latin. (Origin: Norway) • Aslak Bolt’s Bible (Origin: Paris)

14th

15th

• Ms.8º 102 (Psalter, second. prov: Kvikne)

Prague, Archive of the Prague castle Uppsala, University Library

• Psalter. (Origin: Bergen, Munkeliv) • DG 8 II Legendary Olav’s saga. O. N. (Origin: Trondheim)

There are also other, smaller collections with fragments. For a survey of these, see Ommundsen 2006a.

2.1.2 What was once here? Regarding attempts to reconstruct the contents of vanished Anglo-Saxon libraries, Michael Lapidge lists three kinds of evidence which can be used for such a task: surviving inventories of books owned by a particular institution or library; surviving manuscripts attributable to a particular institution; and citations in medieval works

72

2.1. Remains and losses - an estimate drawing on the resources of a particular institutional library (Lapidge 2006, 53). In our case we are not trying to reconstruct the holdings of a library, but the holdings of liturgical books of an entire country. Medieval mainland Norway covered vast stretches of land, but had a relatively small population, perhaps between 300 000 and 450 000 around the year 1300. It is presumed that only a few percent knew how to read and write, and that includes the c. 2000 estimated priests, the members of the aristocracy and some merchants (Bagge 2001, 19). Church inventories Ludvig Holm-Olsen used several inventories in his survey of book-collections and book-owners in Norway (Holm-Olsen 1990, 124-44). The oldest surviving church inventory, from Hålandsdalen, south of Bergen,70 lists eight different liturgical books in 1306, among them a missal, a “legenda” (þessæ legenda), possibly referring to the breviary in which the list was written (Ommundsen 2004), an “aspiciens” referring to an antiphoner or breviary (ibid.), two psalters and a hymn-book. There is a certain evaluation of the books: “æin psaltare forn” (one old Psalter), “æin ymna bok litil” (one small hymn book). Ylmheim (Norum) church in Sogndal had 17 religious or liturgical books in the 1320' s (13 of which are counted in 1321 and four others in 1323), including one sequentiary (sequencio bok). In this inventory the books are not only specified as “old” or “small”, but there is also an evaluation of the quality of the books and their letters: “mœsso bok… cum nota et competenti litera” (missal with notes and competent letters), “aspiciens gott… cum bona litera et nota” (a good “aspiciens” with good letters and notes), ”æinn forn psaltare” (one old Psalter), ”godr gradal per annum” (a good Gradual for the year).71 It is not clear what qualifies as “competenti” or “bona litera”, but one may at least assume that they were easy to read. An inventory of the bishop’s chapel in Bergen in 1408 lists c. twenty liturgical books, among them one sequentiary (æin sequencionarius). Two graduals are described as “vbwnden cum nota” (unbound, with notation) and two missals are 70 71

DN 21, 7. DN 15, 8.

73

Part I: To study manuscript fragments “cum nota badhæ jllafaren” (with notation, both in a bad state), the only Psalter in 72

the list is described as “æin fordærwadh saltare” (one damaged Psalter). It could seem the mass books in the bishop’s chapel were ready for a replacement at this time. It is interesting and worth noting that old books in a bad state were still kept in the church, even though younger books were presumably available. An inventory of the monastic church (not the monastic library) at Kastelle in Konghelle, Bohuslän (at that time Norwegian territory), lists close to twenty liturgical and religious books in 1485; graduals, missals, breviaries, legendaries, antiphoners, a psalter, a canon, two sequentiaries, an ordinal and a few other books.

73

Legal documents Other sources providing evidence of ecclesiastical book collections in medieval Norway are legal documents, including marriage contracts and wills. The earliest documents mentioning books are from June 1298 concerning the theft of the books and ornaments in St. Mary’s church in Stavanger, although the number of books, presumably liturgical, is not mentioned.

74

Private individuals, churches and

monasteries all received books as testamentary donations. Margareta Philippus’ daughter receives among other valuable items a Psalter for her wedding in 1322.75 A Psalter is also mentioned among other possessions in DN 2, 165. A canon of Nidaros bequeathed a breviary to the church, a missal to “Mariu stuku” and a psalter for his “provendo”, i.e. that he in his old age should be taken care of in the monastery, in 76 1343. A document over the division of the inheritance between a brother and sister

from Talgje, Rogaland, in 1366 also contains a collection of eleven books: A law book, with an estimated value of two and a half mark forngild and many books not valued; a Psalter, a book of hours, a book containing Bever' s saga (a translated romance of chivalry) and many other “well written sagas" in the same book, and seven other small books which are both books of hours and saga books.77 A canon in Oslo left his breviary (“the one I read from every day”), a lawbook, a Theodolus (a versified 72

DN 15, 42. DN 14, 158. 74 See for instance DN 4, 26. 75 DN 2, 147. 76 DN 2, 255. 77 DN 4, 457. 73

74

2.1. Remains and losses - an estimate Carolingian primer) and another breviary to friends and relatives in 1368.78 The bishop Botolf of Stavanger donated his private books, most of which seem to have contained canon law, in 1370.79 Property lists There are several lists of the belongings of bishops and archbishops, and books are a natural part of these lists. There is a much used list from c. 1300 with 36 titles on the back of a manuscript in Uppsala: Hos libros possidet b Aquila, possibly referring to 80

bishop Arne Sigurdsson of Bergen (1305-1314).

The list includes 13 different

theological works, 12 grammatical books, and a few books in Old Norse. The bishop of Stavanger bought the book collection of Narve priest of Øyestad (Aust-Agder) in 1338, but what books these were is not mentioned (Holm-Olsen 1990, 130). Archbishop Thrond Gerdarsson (1371-1381) owned the Icelandic saga manuscript Eirspennil: Iste liber est throndonis gerdari canonici nidrosiensis. Archbishop of Nidaros Nikolaus Rusare (d. 1386) brought with him to Denmark belongings of the archbishopric, which were returned to his successor the year after Nikolaus'death. Among other items were a missal in two parts, a breviary, a Bible, legal and administrative books, and a passionale.81 Archbishop Aslak Bolt brought in 1429 c. 20 books with him from Bergen to Nidaros, among these several liturgical books; an antiphonal de sancti with “new notation” (cum nova nota), and a breviary for the whole year “after the Bergen rite” (secundum modum bergis). These books were the 82 archbishop' s private property. During his time in Nidaros he also traded a Bible and

a Decretum for a magnificent illuminated Bible probably written in Paris in the 1260' s, the above-mentioned Aslak Bolt’s Bible.83 Archbishop Henrik Kalteisen made an inventory before leaving Bergen and Norway in 1458, which included a missal and canon law books, and an almost full collection of the works which constituted what

78

DN 4, 475. DN 4, 494. 80 Johnsen 1908, 86. Whether b Aquila really referred to bishop Arne has been disputed by both Stefán Karlson (1979) and Mattias Tveitane (1981) (cf. Holm-Olsen 1990, 131). 81 Original (North German): NRA. Printed in DN I no. 508, DD 4 rk III no. 240, with Norwegian translation in Dybdahl (2002, 112 and 145). 82 Original (parchment): NRA - AM 14.6a. Printed in DN V no. 586, with Norwegian translation in Dybdahl (2002, 114 and 147). 83 The Aslak Bolt Bible is now in the Deichmanske Library in Oslo. 79

75

Part I: To study manuscript fragments was later known as the Corpus Juris Canonici: the Decretum (Gratiani), the Decretales (of Gregory IX), and the (Liber) Sextus (of Boniface VIII) and the Clementines (of Clement V) in one codex. He also had the Sermons of Innocent I, a pastoral on paper and 84

another missal, belonging to the Nidaros Archbishop' s chapel. The Archbishop Erik Valkendorf left his belongings with the Carthusians near Amsterdam on his way to Rome, where he died in 1522. He left a chest with 19 books of various size and value, with wood covers, among them books for the sacred office, and a missal.

85

Estimate based on written evidence The documents above refer to liturgical books in churches and books in private ownership, both liturgical and “library books”. Based on the surviving church inventories it is possible to suggest that a small church would have less than ten books of various kinds, and that a more important church, like a larger parish church, a monastic church or a bishop’s chapel, not to mention a cathedral, could have up to twenty liturgical books. Books of a religious nature could also be in private book collections, not only those of bishops and canons, but also wealthier laypeople, who could use books like psalters or books of hours for private worship. In addition “sagabooks” and lawbooks could be in private possession. Bishops and archbishops could have rather large personal book collections of twenty or more books, which would generally include 2-3 liturgical books, such as a personal missal and a breviary, and several law-books. In addition to the 5 cathedrals and 30 monastic churches in medieval Norway, it has been estimated that there were c. 1200 parish churches.86 Espen Karlsen estimates the number of churches to c. 1300, with an average of about 10 books per church, and suggests a minimum of 13 000 books for the early fourteenth century (Karlsen 2003). Andreas Haug and Gisela Attinger suggest that c. 1200 churches might have possessed a mere 2500-5000 liturgical books (Attinger and Haug 2004, 10). 84

Original: Bonn University Library, No. 326. Printed in Erkebiskop Henrik Kalteisens Kopibok, p. 171 f, with Norwegian translation in Dybdahl (2002, 118 and 153). 85 Original (paper): NRA (Münch. Saml. No. 69). Printed in DN VII no. 604, the testament: DN VII no. 558, with Norwegian translation in Dybdahl (2002, 121 and 156). 86 In his Sammenlignende Fortegnelse over Norges kirkebygninger i Middelalderen og Nutiden Lorenz Dietrichson estimated approximately 1200, of which 30 small chapels (Dietrichson 1888).

76

2.1. Remains and losses - an estimate

In my opinion an average of c. 10 books per church is reasonable, and an estimated number of 10-12 000 liturgical books around the year 1300 is likely. It is also likely that this estimate for 1300 stayed relatively unaltered until the Reformation in 1537. It must be assumed that there was some replacement of liturgical books all through the Middle Ages. Fires haunted towns and churches throughout the Middle Ages, and they must have been responsible for some loss of books. Archbishop Eric Valkendorf’s introductions to Breviarium Nidrosiense (1519) and Missale Nidrosiense (1519) indicates that there was no dramatic increase in book production around the fifteenth and early sixteenth century. In his introduction to the printed missal Valkendorf complains about the shortage of missals in the whole diocese and its province, and that the books which were still used were ancient and hardly legible, written in characters which were out of use and full of errors. He also mentions several secular priests celebrating mass using books of the Cistercian, Dominican or Franciscan orders, although all provincial churches were obliged to imitate the metropolitan church in lessons and chants.87 Although it was in Valkendorf’s interest to express a grave need for printed books to justify their production, his description is to some degree supported by the transmitted material in the NRA. The large number of fragments from twelfth century books indicates that some churches in the fifteenth and sixteenth century kept, and probably also used, missals and breviaries which were several hundred years old. At the time of the Reformation ”the clock stopped” as far as liturgical books were concerned, and they became obsolete. From 1537 the corpus of medieval liturgical manuscripts would only grow smaller and smaller. “Library books” It is more difficult to estimate the corpus of “library books”. C. 1300 there were about 87

Reverendissimus in Christo pater dominus Ericus Valkendorff: dei gratia Archiepiscopus Nidrosiensis et apostolice sedis Legatus: considerans totam diocesim et provinciam suam defectu librorum missalium laborare. Preterea eos qui quotidie in manibus sunt antiquissimos eiusmodi codices inusitatis et desuetis characteribus scriptos vix legibiles esse – non paucis item mendis et erroribus subinde respersos. Atque ob id complures secularium sacerdotum Cisterciensium, predicatorum et minorum ordinum lecturas celebrando observare: quum tamen singule ecclesie provinciales Metropolitanam ecclesiam legendo et cantando iure imitari teneantur. ... (Missale Nidrosiense 1519)

77

Part I: To study manuscript fragments 30 monastic libraries in Norway, and probably several large book collections associated with cathedral chapters and royal administration. In his work on AngloSaxon libraries (which primarily focuses on a period before any major Norwegian library ever existed, but which can be still be useful) Lapidge suggests a “core” of books found in any library of the Middle Ages. This includes classical grammarians and poets, and the patristic authors Gregory,88 Isidore,89 Jerome90 and Augustine,91 but also others, like Cassianus and Eusebius. In addition, a library of a date such as ours would also be likely to have Caesarius’ Sermones and Boethius’ De consolatione philosophiae (Lapidge 2006, 127-128). In the six inventories from private persons and libraries listed from Anglo-Saxon England the number of books varies from 14-60 books. Assuming that a “standard” library in a religious house in Norway c. 1300 would contain at the very least between 10 and 20 books, with grammatical works, some basic reading texts, sermons and homiliaries, commentaries on the psalms and the most important patristic works, like Gregory’s Moralia in Job and Augustine’s De civitate Dei, the c. 30 religious houses would have a total of something between 300 and 600 library books, most of them probably in Latin, but presumably also a considerable number in Old Norse. In addition, the five chapters connected to the bishops’ sees would have libraries of approximately the same size and presumably the same contents. Several books, like the lawbooks, would be required for use in the royal administration or among town officials, presumably mainly lawbooks in Old Norse. Other books, like the novels and saga books, would be sought out by wealthier lay-people, also presumably mainly in Old Norse (including Old Norse translations from Latin). It seems that it is not unreasonable to suggest that at least 500 to 1000 library books of various genres existed in Norway c. 1300.

88

Dialogi, Hom. .xi. in Evangelia, Moralia in Job, Regula pastoralis. De ecclesiasticis officiis, De natura rerum, Etymologiae, Synonyma. 90 Epistulae, Comm. in Evangelium Matthei. 91 De civitate Dei, De trinitate, Enarrationes in Psalmos, Enchiridion, Epistulae, Sermones. 89

78

2.1. Remains and losses - an estimate 2.1.3 How much is lost, and what are the consequences? In the two last chapters I have given a rough survey of both the number of remaining manuscripts, either in the form of codices or fragments, and of the number of manuscripts presumed to exist in Norway around 1300 and at the time of the Reformation. The results can be entered in a table, like this: Table 4: Estimated numbers of books in Norway in the Middle Ages and today: Liturgical books

Library books

Latin

Old Norse

c. 1300

10-12000 books?

500-1000 books?

c. 1537

10-12000 books?

500-1000 books?

today

9 codices

No codices

C. 155 mss in the form

c. 1000 mss in the form

c. 100 mss in the form

of codices or fragments

of fragments

of fragments

Of the liturgical books, it seems that there is evidence of approximately 10% of the total corpus of books which was present at the Reformation. Of these books some were older than 1300, others were produced between 1300 and 1537. Of the library books the numbers are not so different – also here one could say that there is evidence of 10-20% of the books in the collections of fragments. Still, it must be clear that this does not mean that we have 10-20% of the manuscripts or the manuscript material. The transmission of the total number of pages is considerably lower than 10%, and even considerably lower than 1%, since the books are in such a fragmentary state. When it comes to how much less than 1% remains, I will not speculate, as it is clear that the losses are immense. The immensity of the lost material is also illustrated by the low transmission of the liturgy for specific saints. The remaining evidence for the liturgy of important Norwegian saints like Olav, Swithun, Sunniva and Hallvard is, with the exception of Olav, as close to nothing as it is possible to get. Regarding Lux illuxit letabunda, the sequence for St. Olav, it is not unreasonable to assume that this sequence would have been present in virtually every church in Norway. A sequentiary is mentioned in several of the church inventories, and it seems to have been a common book. In 79

Part I: To study manuscript fragments addition several missals and graduals would have included sequences, at least the most important ones. Considering the 1200 parish churches, 30 monastic churches and 5 cathedrals, it would not be unreasonable to assume that Lux illuxit letabunda existed in at least a thousand written copies at the time of the Reformation, and probably more. Still, Lux illuxit is transmitted in only four handwritten Norwegian sources and two Icelandic ones.92 In contrast it is registered in 38 manuscripts in the Swedish fragment collection.93 The English St. Swithun-sequence Psallat ecclesia mater decora is like St. Hallvard’s sequence Lux illuxit... lux est nobis represented only in one 94

Icelandic fragment

– no Norwegian. The liturgy for the saints of Selja and St.

Sunniva, so important in the diocese of Bergen, and celebrated also in the other provinces of Nidaros, has not been identified in one single manuscript fragment. It is of course possible that some unidentified fragments exist, but the chances are not great, as several people with a sharp eye for such things have been looking through the unregistered fragments. This leads us to the sequentiaries. Of the c. 70 Norwegian manuscripts with sequences, approximately 35 are sequentiaries. With the exception of Seqv 1, most of them have only two pieces remaining. In Sweden, for comparison, 243 of the 448 manuscripts with sequences are classified as sequentiaries (Björkvall 2006, 47). While the short booklist in Hålandsdalen does not include a sequentiary, the church inventory at Ylmheim lists one, the inventory of the bishop’s chapel in Bergen lists one, and the inventory of the monastic chuch at Kastelle mentions two. If we infer from this that not all churches, in particular not the smaller churches, had a separate sequentiary, but managed with sequences in missals and graduals, we might make a very careful estimate and suggest that only half the churches owned separate sequentiaries. In that case c. 600 churches would have had a sequentiary, and the evidence of 35 sequentiaries we have today are witnesses to about 5% of them.

92

The Norwegian sources are: Seqv 1 (Lat. fragm. 418), Seqv 13 (Lat. fragm. 986), Seqv 18 (Lat. fragm. 932), Seqv 64 add (Lat. fragm. 1030) (inc. only). The Icelandic sources are Copenhagen AM 98 8° II, ff. 5-8 and Reykjavik, the National museum, no. 3411. All of these have the sequence with lacunae. 93 I thank Gunilla Björkvall for this information, latest confirmed in an e-mail of 20.10.06. 94 The opening lines of Psallat ecclesia mater decora is found in Reykjavik, the National Library, fragm. 19, 1v (available online at http://ismus.musik.is). Lux illuxit... lux est nobis is in Copenhagen AM 241b fol. IV, f.1

80

2.1. Remains and losses - an estimate Again, as only a few pieces remain from most of these 35 sequentiaries, the surviving number of “sequentiary pages” is well below 1%. The consequence of such a poor survival rate is first and foremost that one can never draw a conclusion based on the lack of evidence (the socalled argumentum ad ignorantiam). That something is not found among Norwegian manuscripts or manuscript fragments, can not be taken as evidence that it never existed. So, for instance, a discussion starting with the question “why was there never written a sequence for St. Sunniva?” would be invalid, because a sequence could very well at some point have existed – it has just not survived, or been identified among the fragments. When the group of manuscripts or fragments is so small compared to what once was, the statistical results regarding date, origin, book genre and so on, will not be valid for a wider group than this particular corpus of manuscripts. This study is a contribution in terms of specific evidence, but the results can only to a very limited degree be generalised. Broader statistical results can only be achieved through more studies of this kind.

81

Part I: To study manuscript fragments

2.2. Are the liturgical manuscripts representative? I assert that it is possible to study book and scribal culture through liturgical fragments, but to what extent can liturgical books be seen as being representatives for a book and scribal culture? One might say that since the liturgical books are more or less “mechanically copied”, and leave little room for creativity so far as the contents are concerned, library books would serve this purpose better. This will be discussed below. The question above can be divided in two parts: 1) Can the liturgical manuscripts be seen as good representatives for a book culture? 2) Are the fragments we have today from the liturgical manuscripts representative of the liturgical manuscripts which once existed?

2.2.1. Liturgical manuscripts as books “Book and scribal culture” is an inclusive term. While “book culture” can embrace various aspects of literary study and literary production, “scribal culture” is connected to the art of copying and writing a book, regardless of the “originality” of the contents. My point of departure is that liturgical books to a certain degree can be used as representatives for both book and scribal culture. It might seem obvious that liturgical books will be of limited use if the purpose is to study a growing level of education and the intellectual life of a learned elite. Still, liturgical books contain original compositions and texts which testify to a high level of Latinity, and the liturgy reveals a general ability among the men of the church to participate actively in the selection and production of material for their local liturgical use. But that aside, one may ask if the European influence found in the liturgical books is representative of what we would find in other book genres, and if the liturgical manuscripts are representative sources to the growth of a local book production. This will be discussed below. Christianity is the religion of not one, but many books, and the arrival of writing is closely connected to the arrival of Christianity. Liturgical books were probably among the first to be brought to Norway. Since it was close to impossible to celebrate

82

2.2. Are the liturgical manuscripts representative? Mass without the proper Mass books, there must have been a large number of such books at an early date. The estimates in the last chapter indicate that liturgical books far outnumbered other genres in the Middle Ages, and with evidence of a thousand litugical books in modern collections, they far outnumber other surviving manuscript material from the Middle Ages today as well. So, liturgical books were not only the first books to be brought from abroad and the first to be locally produced, they were also the majority of the books in Norway in the Middle Ages and are the majority of the manuscripts testified to among the remaining sources today. In short, the liturgical books not only can, but ought to, be studied as a vital part of Norwegian book and scribal culture.

2.2.2. Fragments as manuscripts Now to the question of the fragments. Are the liturgical books we have evidence for today representative of the ones which once existed? Medieval manuscripts came in many different shapes and forms, depending on the need they were written to fulfill, how much money the commisioner had, and how good the scribes were. A large part of the manuscripts used were “every-day” manuscripts, made to cover basic needs, either for private study, sermons or for liturgical celebration. These were not necessarily written well or given much or any decoration. Some books, on the other hand, were so well written and so lavishly decorated, that they may be characterised as precious works of art. Among preserved codices many are of a high quality and as such they may not be representative of the general standard and form of the books which were produced and used in larger numbers. Of the nine remaining liturgical codices from the Norwegian Middle Ages at least three psalters and one Bible are lavishly illuminated, and of a remarkably high quality. The other five books are of a more basic kind. The English Gospelbook is of a general high quality, but with modest initials (and in one instance a major initial is not entered). The Kvikne psalter is a fine example of local book-production, but also decorated with modest coloured initials. The three manuals do seem to be typical “everyday” manuscripts. That almost half of the remaining Latin manuscripts are precious examples of medieval

83

Part I: To study manuscript fragments art stands in clear contrast to the fragments, where a large number seems to be rather plain and “ordinary” books, although some of good quality. One may imagine that the bailiffs would shy away from cutting up the finest books, and that this is the reason why the standard of the fragments are relatively modest. Some fine examples of illuminated psalters among the fragments do indicate that the people cutting the books did not only choose the plain manuscripts, but used what was available to them, regardless if the book was large or small, or if it was a plain old massbook, or a psalter with gilded and decorated initials. The reuse of parchment seems to be a simple acknowledgment of parchment as a solid material, without any interest or conscious selection from the people using them. For the people of the sixteenth and seventeenth century there was no basic difference if manuscript leaves were reused to measure powder for canons or fireworks (cf Pettersen 2003) – a treatment they would not survive - or to strenghten the binding of a tax account. The difference for us is, of course, that one group of reused fragments happens to be transmitted to us by chance, the other perished. The apparent sixteenth and seventeenth century indifference to the medieval books has the “positive aspect” that they were randomly selected, and for that reason we may feel more confident that the manuscripts we have evidence of may be considered representative for the selection of liturgical manuscripts in the Middle Ages. There is one problem, and that is that not all the tax accounts with fragments in the bindings have survived. On the contrary, large numbers of them have been destroyed. It seems that the larger part of the accounts older than 1610 were discarded in a clear-out in the Danish National Archives in the eighteenth century (Pettersen 2003, 50). This means that most of the surviving fragments were used for the binding of accounts between 1610 and 1650, and these probably represent the last pieces of the surviving medieval manuscripts after the Reformation. Before 1610 manuscripts were probably to a larger degree used as covers, not merely as small pieces to hold the seams of the bindings. When using a leaf as a cover, a large format book would be preferred, i.e. a bifolium from a quarto, or a leaf from a

84

2.2. Are the liturgical manuscripts representative? book in folio format, not a book in octavo format. Of the 69 manuscripts studied here, 22 are in octavo format, 27 in quarto, 10 in folio and 10 unknown. A second difference between someone using bindings in the late sixteenth century as opposed to after 1610, is that while liturgical manuscripts were “everywhere” in the decades after the Reformation, they were a scarce commodity in the early seventeenth century. In 1622 king Christian IV issued a rescript ordering Danish and Norwegian bishops to collect and submit old books to the Danish chancellor in Copenhagen (cf. Pettersen 2003, 49). One may imagine that to find material for bindings one would in the seventeenth century to a larger degree move out of the larger towns to find remaining manuscripts still kept in old parish churches. So it seems that apart from the efforts made to make the parchment last longer, i.e. cutting it into smaller pieces, two things may be different in the remaining material compared to the manuscripts in the earlier, discarded, bindings: The format (more octaves) and the places they were collected (possibly outside the larger centres). Although the format of the remaining fragments may not be representative of the earlier bindings, it might not be so bad compared with the original books. A large part of the medieval books were probably octaves, so the bindings may in fact be more representative of the format of medieval books now than if there had been no shortage at all (cf. the Swedish material discussed below). Regarding the assumed collection of the books out of the towns, this may also not have a huge effect in terms of how the material should be evaluated, as the books in parish churches were probably provided by scribal centres in towns and monasteries. The Norwegian parchment fragments differ from the Swedish fragment material both in extent and format. Not only do the Swedish National Archives have a considerably larger number of fragments (c. 22 700 single fragments compared to c. 5-6000 single fragments in Norway), but the fragments are mainly in the form of bifolia of quarto size. In Sweden bifolia continued to be used as wrapping material throughout the seventeenth century, and there seems to have been no shortage of material. While Norwegian fragments are from books in octavo format (which are

85

Part I: To study manuscript fragments then cut into even smaller pieces), in Sweden octavo books do not seem to have been seen fit to use as binding material. Another difference between the two collections is that Norway has more of the oldest material, i.e. tenth and eleventh century. This may be due to the earlier arrival of Christianity in Norway, perhaps in combination with the old liturgical books being longer in use. While there are considerable losses from Norwegian parchment fragments due to the discarding of old accounts and general losses, Jan Brunius has estimated a general loss of 50 % in the Swedish material (Brunius 2005).

2.2.3. The scribes That scribes writing liturgical books also wrote other types of material is attested to by scribes writing both Latin liturgical books and works in Old Norse. One missal studied here (Seqv 39b add) was written by the same scribe (or a very similar hand) as the scribe of the Old Norwegian Homily book (AM 619 4°). One scribe, called the St. Olav scribe, wrote several liturgical manuscripts, and also the Legendary Olav’s saga (Uppsala, MS De la Gardie 8 II) and an Old Norse law book. The scribe of the famous work Konungsskuggsjá or the king’s mirror (AM 243 b

fol.), also wrote a

lectionary (Gjerløw 1968, 34-38). It has sometimes been suggested that a scribe writing a liturgical book in Latin would generally write in a more formal fashion than when he was writing a “secular” piece of literature. This is not the case for any of these three scribes. The general aspect of their hands is the same and easily recognisable, even in spite of the different letterforms used for the vernacular. It is always possible that some scribes would write more formally when writing Latin, but so far there is nothing in the Norwegian material to suggest this. Therefore liturgical manuscripts appear to be representative of the local scribes’ general style of writing. In conclusion it seems clear that the liturgical fragments are not only representative of book import and for local production, but also as far as the general standard of medieval books is concerned. Thus, they form one of the most important sources for

86

2.2. Are the liturgical manuscripts representative? the study not only for the arrival of writing, but also for the local development of a book and scribal culture.

87

Part I: To study manuscript fragments

3. Approach and method While the last chapters were devoted to the past work on fragments, and the assessment of the volume of available manuscript material, here I will briefly sketch different ways of selecting a manageable corpus among the fragment material, the Latin fragments in particular, and a method to extract the most information from it. The general purpose of this method is to recognise European influence, and see ways to group separate fragments so that they may be easier to relate to a historical context. While this study is basically a genre study where the selected corpus has the presence of sequences in common, it is clear that there are alternative ways to select a corpus for fragment or manuscript studies, which may prove fruitful.

3.1. Selection Considering that more than 99% of the manuscript material which existed in Norway in the Middle Ages is lost, it is ironic that it is so difficult to get started on studies of the remaining parts of the manuscripts. Although one may agree that the fragmented remains of 1200 books in the NRA are the most important and promising witnesses to medieval book culture, it is still not evident where the best place is to start. One study can not, no matter how extensive, include 5-6000 single fragments. There must be something governing the selection, making the number of fragments or “reconstructed” or “once-complete” manuscripts manageable.

3.1.1. Genre study This current genre-study, which involves c. 70 units, i.e. the once-complete manuscripts, the two codices and the parchment roll, is comparable in size with another genre-study, namely Lilli Gjerløw’s Antiphonarium Nidrosiensis ecclesiae (1979) which also contains c. 70 analysed manuscripts, mainly antiphoners, but also breviaries, breviary-missals and breviary-lectionaries. A genre study, either governed by book genre or elements in the contents, is a practical approach at the current stage of knowledge. The fragments in the NRA have traditionally been listed with respect

88

3.1. Selection to genre, and the lists in Lilli Gjerløw’s catalogue and Espen Karlsen’s additions to these lists is a good point of departure. Several genres lend themselves well to such studies, like patristic or homiletic texts, psalters, or other types of liturgical books. A number of between 50 and 100 manuscript units can be manageable in one study, but to include much more would be difficult. The primary purpose may be to study script, decorations, contents, or other aspects of the medieval book and book culture. Liturgical studies have in several cases sought out and selected material which could supply information for the celebration of one particular feast, like St. Olav on 29 July (Reiss 1912, Østrem 2001) or the Holy Blood in September (Attinger and Haug 2004). In the case of the latter there is only one witness to the celebration of the feast. This type of study greatly benefits from a searchable database, as it is easier to search for a feast than scribal characteristics. Different selections of materials could for instance be: - one single manuscript (or groups of related manuscripts). - the products of one scribe (or groups of related scribes) - the products of one scriptorium.

3.1.2. Manuscripts, scribes or scriptoria Just as one single manuscript may be chosen for a study when complete, an incomplete manuscript in multiple fragments could be just as valid a topic. Apt for such a study would be once-complete manuscripts with a high number of fragments still remaining. The purpose could be to “reconstruct” the manuscript and analyse its contents, discuss origin and perhaps find related manuscripts in the form of codices in other collections or libraries.95 In such cases it could be an advantage if the 95

Several once-complete manuscripts in the NRA have c. twenty fragments or more left of the book, like Lat. fragm. 96/97 (psalter, 29 fragments), Lat. fragm. 108 (psalter, 23 fragments), Lat. fragm. 537 (lectionary, 46 fragments), Lat. fragm. 681 (antiphoner, 20 fragments + 2 in Stavanger State Archive), Lat. fragm. 787 (lectionary, 18 fragments), Lat. fragm. 1106 (gradual, 33 fragments). These are just randomly picked examples to show that not all once-complete manuscripts are preserved only through a few pieces. Some items in this study also have more than twenty pieces, like Lat. fragm. 418 (sequentiary, 46 fragments) and 780 (antiphoner, 25 fragments). Several of the mentioned items above are from high quality imported manuscripts, like the psalters Lat. fragm. 96/97 and 108, which were

89

Part I: To study manuscript fragments manuscript was imported and one could look outside Norway to identify related manuscripts.

96

Regarding the studies of the products of one scribe, the choices are limited and related to local scribes unless foreign scribes are recognised in the material. So far at least three scribes are identified as writing in both Latin and Old Norse (cf. Gjerløw 1968, 34-35), two of which are treated in this study (Seqv 38, Seqv 45b and Seqv 64). One more scribe in this study may be identified in two or three liturgical books (cf. Seqv 51 add). As the work with the fragment collection continues, more scribes will probably be identified. To study the products of a scriptorium is premature for our material – it is possible to get there, but it may take some time.97 A possible approach for the time being is to single out fragments with similarities in script or decoration, which may be explored as having a common background, like Michael Gullick has done in two recent studies of Scandinavian material (Sweden and Norway) (Gullick 2005, and forthcoming). Here particular features in script or initials are compared in different manuscripts. While Gjerløw has identified fragments from Benedictine, Premonstratensian and Augustinian manuscripts (Gjerløw 1968, 82-4), Gullick has also made a list of Cistercian manuscripts, based on the presence of particular punctuation mark (punctus flexus). The approach would be the same if looking for books from one particular place, like Lund or other centres – to single out the significant features and look for them among the fragments. It is clear that such a search among thousands of fragments depends much on luck in the current state of knowledge and availability of the fragments. There is, however, one other approach that can be followed in looking for a local scribe or scriptorium, namely to look at fragments used to bind accounts from the probably made in England s. xiii¹, and the antiphoner Lat. fragm. 681, which has a French origin (s. xiv¹). 96 One type of imported book one would probably find equivalents to in other libraries would be the Lat. fragm. 47, Petrus Lombardus’ Commentarius in psalmos Davidicos, a high quality thirteenthcentury glossed manuscript. 97 An example of this type of study is Parkes on Wearmouth-Jarrow (Parkes 1982).

90

3.1. Selection same area. Although there are many considerations to be made regarding bindings and secondary provenance, nevertheless this has proven to be successful in the past (cf. Seqv 1, Seqv 38/Br-Mi 3). This could be an interesting approach in the case of Bratsberg/Bamble, for instance, since the same scribe has been identified on three different leaves used to bind accounts from this fief. It would be interesting to see not only if more fragments with the same hand appear, but also if other scribes with the same training or features occur. Then one would have indications of a scriptorium, and probably a scriptorium near by (like Gimsø). Such an approach would be easier in a place like Bratsberg with not so many scribal centres as the larger towns, where there would be more suppliers of liturgical books.

3.1.3. Manuscripts within a timeframe As the fragments in a database will be dated, it will also be possible to search for the oldest material, for manuscripts older than 1100, for instance, or for the first half of the twelfth century, and make a study across genres. The earliest material is particularly interesting as evidence for the making of a new scribal culture, as evidence of early imported exemplars or early local copies. These could reveal more about the constellation of foreign and local scribes in the first scriptoria and the quality and nature of the very first books. In this study, after getting an overview over the manuscript corpus, I chose to focus on fragments from manuscripts with an origin within the timeframe c. 1100-1300, and selected manuscripts from this period for closer analysis. Now that approaches to the fragment material have been discussed, the topic of the following chapter will be ways to date and estimate origin.

91

Part I: To study manuscript fragments

3.2. Searching for European centres in Norwegian fragments For the dating of a manuscript the guidelines are clearer and more numerous than when it comes to suggesting an origin. This is especially challenging when nothing is known about the use or historical context of the manuscripts, and only small pieces of the contents are available. In the opening words to this study I stated that the liturgical fragments could be used to study a growing manuscript culture and its European influences. In this section I will discuss how these influences can be identified and traced in specific fragments, and outline some guidelines which I follow in the analysis, and which may be applicable for future studies of manuscripts or manuscript fragments. Before moving on to the indicators of origin, a few words should first be said about dating a manuscript, particularly for the period mainly dealt with here, namely the transitional period from Pre- or Protogothic to Gothic style script. There are several good guides for dating manuscripts (Bischoff 1990, Ker 1960a, Derolez 2003).98 One may look at features in the script, the lay-out, the ruling or the decoration. When several things are considered and compared, one may estimate a date. Dates will always be tentative, as there are factors that may influence the appearance of the script: Medieval scribes were people, and did not necessarily do things in the ways that modern textbooks say they ought to do them. One scribe could write in the same style for a life-time, while another could keep up to date on the latest developments in Europe, which is why it is difficult to specify a date closer than the span of one generation. In some cases allowance should be made for some delay for changes in script to reach the northernmost parts of Europe. On the other hand, one should not automatically assume that all Scandinavian scribes were lagging behind. Another caution concerns the liturgical books with staves for musical notation: the script adjusted to such a lay-out may look rounder and flatter than a prose page, deceiving the palaeographer trying to date the script.

98

Books with plates of dated manuscripts, like Thomson 1969, are useful. Plates of dated manuscript material are also available for Norway (Hødnebø 1960), Iceland (Benediktsson 1965) and Denmark (Kroman 1951).

92

3.2. Searching for European centres in Norwegian fragments Evaluation will ultimately be based on personal judgement, but it is important to present some of the guidelines to evaluate a given manuscript or manuscript fragment. After a corpus is selected, the following steps will be to: 1) Divide the locally produced manuscript material from the imported material so far as possible. 2) If imported, specify whether the manuscript was written in the Germanspeaking parts of Europe, the Anglo-French area, or possibly southern Europe or Italy. 3) If locally produced, determine the influence from a European geographical area on the hand(s) in a manuscript.

3.2.1. Dividing the local from the imported Regarding the first point a “handlist” of indicators of local production was discussed at the CMS workshop The beginnings of a Nordic Scribal Culture, ca 1050-1300 in October 2005 and presented in the report from this workshop (Ommundsen 2006a, 40-43). The points in the list of indicators were based mainly on the work of Lilli Gjerløw and Espen Karlsen, and the observations at the first workshop in the NRA in August 2003. The making of “check-lists” can seem like a naive approach to a complex and difficult field. Still, it is important to start with the specific signs on a manuscript page to make a huge and diverse source material manageable. The following list presents four main categories, and compared to the list in Ommundsen 2006a, the order has been changed so that the most “secure” indicators are listed first: 1. Contents -

the presence of local saints or feasts

-

liturgy identifiable to the local rite

2. Contextual issues -

same scribe writing in the vernacular and in Latin

-

same scribe appearing in fragments from more than one manuscript

3. Links to the Vernacular language or script

93

Part I: To study manuscript fragments -

endings in ‘a’ where Latin would have ‘o’: lectia, postcommonia, prefatia (presumably influenced from Old Norse pronouncination, as opposed to the Latin lectio, postcommunio, prefatio)

-

‘au’-ligature (from Old Norse script)

-

‘ae’-ligature (from Old Norse script)99

-

small cap ‘R’ in larger numbers, and in the beginning and middle of words (not just in endings and nomina sacra) and small cap ‘H’ (not in nomina sacra) (from Old Norse script)

4. Execution -

awkwardness (inconsistency in execution, lack of training)

-

old fashioned traits (like the e caudata in English type script after 1200)

-

mixture

of

Anglo-Saxon/Anglo-Norman

and

German

influence

in

combination with mediocre quality (at least for the eleventh and early twelfth century). Even with the help of this check-list it is still difficult to determine with certainty whether a manuscript was written in Norway/Iceland or not. When pointing to poor quality as an indicator of Norwegian/Icelandic manuscripts, two questions are natural to ask: 1) Were no manuscripts of poor quality produced in the central areas of Europe, like England, France, Germany or Italy? 2) Were no manuscripts of good quality produced in the northern periphery of Europe? The answer to both questions is “yes”. Manuscripts for private use, notebooks and so on, with a homemade appearance have survived from all over Europe. In the “official” ecclesiastical context, however, we may assume that the constant demand of liturgical books would give room for more effective scribal “machinery” connected to the production of these books, and that the supply of good quality 99

This was not part of the original list, but this feature, which is not normally found in Latin manuscripts from the twelfth and thirteenth century, may be ascribed to an influence from Old Norse or the other Scandinavian vernaculars. The ‘æ’-ligature occurs more than once in the Norwegian sequence manuscripts (mainly as the result of misunderstandings – cf. Seqv 2 and Seqv 53 add). Although the æ-ligature is also part of the English vernacular alphabet, it does not seem to be found in English manuscripts in Latin at this time.

94

3.2. Searching for European centres in Norwegian fragments material was better than in a more remote area. With respect to the second question, the larger Norwegian scribal centres were probably able to produce books of high quality.100 How many of the scribal features listed above are common for all parts of Scandinavia? Books assigned to Norway are often done so simply because their presence in Norway makes a Norwegian origin more likely than an Icelandic, Swedish or Danish origin. This is the case also in this study. Even though we do know that there was a large number of Icelandic manuscripts in Norway in the Middle Ages (cf. Karlson 1979), we still do not know if this also applies to liturgical manuscripts. In other words, we do not known enough of what characterises the scribes of the different Nordic countries at different periods – what they have in common, and what may enable us to distinguish between them. The tools to recognise features from different European regions, not only in imported manuscripts, but also in those locally produced, is the topic of the next chapter.

3.2.2. Localising regional European traits Moving southwards from Scandinavia and Norway, the other parts of northern Europe will for the sake of simplicity be divided into three main areas: England, Germany (the German-speaking area) and France (the French-speaking area). As for the local manuscripts, the first step in localising an imported manuscript is to see if the contents can reveal something about where it was made, either if it contains a particular local saint or feast, or follows a specific rite. By Felix Heinzer this is referred to as the “Basistext-Methode”, in which is also included the use of medieval calendars (Heinzer 1984, 89-99). As an example of an unused method for assigning an origin, Heinzer also mentioned the use of the listed sources in Analecta Hymnica (ibid., 103). If the contents do not provide any clues, the evaluation of origin depends on palaeography and codicology.

100

One example of such a book may be NRA, Lat. fragm. 489 (old number) (Seqv 33b add/Mi 107).

95

Part I: To study manuscript fragments An important remark of caution is that scribes were mobile (cf. Karlsson 1999, 146). Palaeography can only identify where the scribe received his or her training, not where the person later worked (cf. Powitz 1976, 132). Here codicology comes in with additional evaluation of the parchment, the decoration, and the binding (or in the case of fragments: possible traces of holes from the original binding). Still, some scenarios will be difficult if not impossible to trace. The product of an English scribe working at one of the larger scriptoria in Oslo, Trondheim or Bergen, with a supply of proper parchment and good quality ink, could, with the applied method, be assigned to England, unless we are fortunate enough to have elements in the contents pointing to the liturgy of Nidaros. Likewise a manuscript written in a small scriptorium in Germany, by someone who had moved around a lot and developed a “composite” hand combining elements from several European styles could be assigned to Norway. The ways of recognizing regional scribal features have not been studied in the same way as dating, and is treated with much more caution. Derolez’s Gothic bookhands does exhibit examples of different executions of letters both in Pre-/Protogothic and Gothic script, but is careful when it comes to assigning these executions to regions. In the subchapter “regional differences” he emphasises how heterogenic the Pregothic script is, and how difficult it is to find characteristics special to the European countries (Derolez 2003, 71). When he refers to an earlier attempt to separate the European regions, it is more as a curiosity than anything else, since descriptions like “in Italy the script is larger and more beautiful than in other countries” or “script in France is close to English Pregothic, but heavier and rounder, and sometimes angular” (Derolez 2003, 71, with ref. to Battelli) is of very limited use in practical palaeography. Derolez’s main focus in terms of regional differences is the contrasts between the northern and southern textualis. As most of the manuscripts and fragments dealt with in Scandinavia are northern European, this is only partially relevant for

96

3.2. Searching for European centres in Norwegian fragments someone wanting to separate the regions of northern European manuscripts.101 To my knowledge no-one has put up a “checklist” for the different regions of northern Europe, and probably with good reason. Medieval life and book history was more complex than a modern “system” can trace. Still, it is necessary to have a point of departure for expressing an opinion, and to make a foundation from which people can evaluate for themselves the arguments regarding each manuscript. I propose one method here, with the help of Ker, Derolez, Thomson and the experienced palaeographers in the NRA workshops, who have mentioned examples of how specific features can be applied in search for origin. It must be clear from the outset that none of the answers we get out of such methods are absolute. While the traditional Norwegian approach to a palaeographical description has been to describe the executions of the alphabet, starting with a, b and the subsequent letters of the alphabet, I will suggest a different first step and state that there are a few significant letters and signs worth noticing more than others, without going through the whole alphabet. The method used in this study, and mostly relevant for twelfth and thirteenth century manuscripts, has been to evaluate a combination of six different features in the script. None of these features should be used alone, but in combination with the others. The six features are the shapes and ductus of the letters a and g, the ampersand used for “et” (&), the punctuation mark punctus elevatus (.’), how the minims meet the baseline and how the smaller initials are made. 1) The letter a: The two-compartment a of Protogothic script and the later textualis has several variant renderings. One, the socalled “trailing-headed” a is considered to be typical for England, the other, which has a different ductus, is seen in manuscript from French-speaking and some of the German-speaking parts of Europe. a) Trailing-headed a:102

101

There are examples of Italian law-manuscripts in Scandinavia, cf. Brunius 2005. The trailing-headed a below is taken from Ker 1960a Pl. 19, and turned into a bitmap image. See also Thomson 1969, Pl. 84, 85, 87 and 88.

102

97

Part I: To study manuscript fragments N. R. Ker identifies a form of the letter a which he considers to be typical for English manuscripts of the twelfth century, namely the “trailing-headed” a (Ker 1960a, 36). The characteristic of this a is that the headstroke projects farther to the left than the lower lobe. While the earliest examples of this were with a in initial position, it was during the twelfth century used also in medial and final position. This execution of the a was not used by everyone, and is often also used in combination with “normal” a. The ductus of this a often leads to the letter extending somewhat above the x-line 103

(cf. ibid).

b) “Flat-headed” a:104 The “flat-headed” a, if one may call it that, is mainly seen in manuscripts from the French- or German-speaking parts of Europe. In England it is used in display-script (i.e. only rubrics or the first words in an opening sentence), but not in running text.

105

This particular type of a has another ductus than the “normal” a. The left part of the letter sometimes looks like it is written like a reversed s. The result is a triangular lobe with the starting point high up on the stem of the a. The high starting-point of the lobe of the a also seems to be a feature of some French and some German manuscripts also when the ductus is “normal”. In England the lobe of the a generally starts lower down on the stem. In most cases there is nothing unusual about the letter a, and it does not point in any particular direction. If, however, the a is trailing-headed this could point in the direction of England. If the a on the other hand is socalled “flat-headed” this would point away from an English origin, and towards French- or German-speaking parts of Europe.

103

A tall a can also occur in French and German manuscripts, see for instance Thomson 1969, Pl. 39. The image of the flat-headed a is taken from an exhibition catalogue no. 58 for the Herzog August Bibliothek in Wolfenbüttel (Ganz, Härtel, and Milde 1989), 147 (ms Cod. Guelf. 224 Gud. lat.: Sextus Propertius, c. 1200, northern France), and turned into a bitmap image. For other examples of the flatheaded a, see Thomson 1969, Pl. 35 (in a manuscript written in St. Lambrecht, Austria 1216). 105 It is for instance used in display-script in the English missal from s. xii/xiii in Stavanger State Archive (fragm. 1-3), which is registered in Lilli Gjerløw’s catalogue as Mi 35. 104

98

3.2. Searching for European centres in Norwegian fragments 2) The letter g: The letter g as it is often found in English and French manuscripts has two lobes, the lower sometimes closed with a hairline, sometimes left open. The ductus of the lower lobe often starts to the left side of the upper lobe, making the body of the letter almost go in an s-like shape:

106

In several manuscripts, the lower lobe can end in a horisontal line or turn downwards to the left, and be closed with a hairline (or simply left open).

107

a) Eight-shaped g: 108

In England the g of the later half of the twelfth century can look like the number 8, as shown above (Ker 1960a, 35; Derolez 2003, 62). b) Straight-backed g:

109

A type of g with another ductus is not found in England, but in the German-speaking parts of Europe and in parts of France. Here the lower lobe starts in a continuous line from the right side of the upper lobe. The shape of the g can be a good indication, but must be used in combination with other signs. The most certain conclusion one can draw based on the shape of a g is that “if the letter g is straight-backed, the scribe is not English”. In such cases the manuscript is probably from Germany or parts of France (or other parts of southern Europe). If the lower lobe or bow of the g starts with a new stroke, the manuscript is most likely English or French. There are some German manuscripts with the same

106

The g below is taken from Ker 1960a, Pl. 19 and turned into a bitmap image. The g is taken from Thomson 1969, Pl. 12, and turned into a bitmap image. 108 The eight-shaped g is taken from Derolez 2003, Pl. 12, and turned into a bitmap image. 109 The straight-backed g is taken from Thomson 1969, Pl. 31, and turned into a bitmap image. 107

99

Part I: To study manuscript fragments ductus of the g, but in these cases there is often some other feature revealing at least that the origin is not English.

110

3) The ampersand (&): According to Derolez the ampersand is the graph with greatest variability in the twelfth century (Derolez 2003, 66). Although there are few absolutes also here, and different variations are found in different places, there seem to be tendencies worth noting. a) “English-looking” ampersand:111

The ampersand mainly used in England is well balanced and upright. The ductus of the final stroke (on the right side) goes downward to the left with an approach-stroke to the left of the line (cf. Derolez 2003, 66). It does in several English manuscripts “tilt” more to the right than in the example shown here, but the head of the ampersand is usually smaller than the body in the English manuscripts. If the ampersand tilts to the right, and “body” and “head” are of equal size this points away from England.112 It has been said that the “typical” English ampersand resembles a person sitting upright, reading a book.113 The basic ductus as shown here is also found in France and Germany, but usually the ampersand will be more forward-leaning outside England. b) “Continental” ampersands:114 or with the “head” resting on the final stroke:

110

See for instance Derolez 2003, Pl. 21, and Thomson 1969, Pl. 41. In the first case the hairline on the r points to Germany and away from England. In the second case there is a conspicuous lack of serifs or “clubs” when the minims meet the baseline. The g can also be said to be eight-shaped in the manuscript from Italy in Derolez 2003, Pl. 7, but also here there will be signs showing that the manuscript is not English, like the shape of the g or the ampersand. 111 The “English-looking” ampersand is found in Ker 1960a, Pl. 19. 112 There are examples of “un-English” ampersands in manuscripts copied in England, see for instance Parkes 1993, Pl. 51, 250. Notice here both the shape of the ampersand, bottom line, and the “flatheaded” a 5 lines from the bottom (mid line). 113 I thank Teresa Webber for quoting this vivid image for me. 114 The “Continental” ampersands used as examples are found in Thomson 1969, Pl. 1 (France) and 34 (Germany).

100

3.2. Searching for European centres in Norwegian fragments Here the final stroke is made upwards from left to right, terminating in an endstroke to the right of or below the final stroke (cf. Derolez 2003, 67). The ampersands exhibit varied degrees of “tilting”. The forward lean probably also to some degree governs the ductus of the final stroke. To continue the image above, some of the ampersands found in parts of Germany or France (although the basic “English” type is also common here) looks like different stages of the reader falling asleep, about to drop the book. 4) The punctus elevatus (.’): The punctus elevatus indicates a major medial pause, and can be a useful sign when looking for the origin of a manuscript as it has several different forms (of which only a few will be looked at here). Its use, development and different renderings are explained in the very useful book Pause and effect (Parkes 1993, 42-43). The punctus elevatus in English or French manuscripts often has a cup-shaped “tick” or a straighter tick with an entry-stroke from the left. In French manuscripts this entrystroke can be quite sharp,115 the tick can simply be an oblique line with no entrystroke (.’),116 or the tick can be traced in a straight stroke upwards with an exit-stroke going down to the right.117 In some southern German manuscripts is found a punctus elevatus with the entry-stroke from the high left, resembling a number seven. This should not be confused with the punctus flexus (“seven and point”) especially connected with Cistercian and Carthusian practice.118 A punctus flexus would be used in combination with punctus and punctus elevatus, and would normally have a longer horisontal head-stroke. a) Punctus elevatus most common in English or French manuscripts:119 or These shapes or varitations of them can also occur in manuscripts from the Germanspeaking parts of Europe. 115

See for example Thomson 1969, Pl. 5. See for example Thomson 1969, Pl. 7. 117 See for example Derolez 2003, Pl. 3. 118 Cf. Ker 1960a, 47-48 and 58-59, Parkes 1993, 306. 119 The example of an English cupshaped punctus elevatus is taken from Ker 1960a, Pl. 19. For an example of this punctus elevatus in France, see Thomson 1969, Pl. 3. 116

101

Part I: To study manuscript fragments

b) A punctus elevatus occurring in manuscripts from southern Germany or Austria:120

5) How the minims meet the base-line. Number five in this list of noticeable traits is the way the minims meet the baseline. This feature not only says something about the degree of formality in the script, but can also indicate where a scribe was trained. In general most scribes simply end the minims with a slight turn upward to the right. There are, however, exceptions to this, and the following points can be suggested: a) If there are no serifs or “clubs” on the minims of a twelfth or thirteenth century manuscript, or if other minims turn to the right or have a serif, but the mid-stroke of the m ends with no turn or serif, this points to the German-speaking parts of Europe.

121

b) If the minim ends in a very sharp and angular serif upward to the right, this points to the French-speaking parts of Europe. c) If the minims end in a flat hairline or wedge with a horisontal base – a type of formal script referred to as textus praescissus – this could point to England, as this type of script was particularly favoured there. It was also used in Germany, but only sparingly in France (Derolez 2003, 76). 6) The smaller initials Michael Gullick makes a distinction between the pen-drawn and the pen-written smaller initials for twelfth century manuscripts. While he sees the first type as typical for England, the pen-written initials are found mainly in Continental Europe. Parkes treats the function of the smaller initials, but do not make regional distinctions 120

See for example Parkes 1993, Pl. 66 (Switzerland), Thomson, Pl. 33-35 (Southern Germany and Austria). This feature may have extended further north, see example from the Low Countries in Derolez 2003, Pl. 5. 121 For the simple midstroke of the m, see for instance Derolez 2003, Pl. 10, Thomson 1969, Pl. 32.

102

3.2. Searching for European centres in Norwegian fragments (Parkes 1993, 43). The English pen-drawn smaller initials are sometimes drawn in black ink, sometimes in colour. The pen-written initials are often touched with colour, especially red. In some cases one may get quite close to an origin when using these features in combination, in other cases we can only see what something is not. But to be able to say “this is not an English manuscript” is also a kind of progress. After the evaluation of these different features, several other smaller features could be considered, like the shape of the r (regarding the “horned” r used in Germany, see Derolez 2003, 83, and Thomson 1969, Pl. 37), the dotting of the y in the English manuscripts, crossing of the x or z or some of the abbreviation-signs. After a while, using such specific traits as mentioned here, one may be able to develop an eye for the “general appearance” of the script of a region. I do believe that the first step to make “das Auge sehen” requires a search among details. As mentioned in chapter 1.2., not only palaeography can be used to trace origin. There are also the codicological aspects, like the parchment, the holes from the original binding, the decoration and so on, which also has a potential in the search for origin, and which may say more about the origin of the manuscript, rather than the origin of the scribe. The notation can also be an additional tool for arguing for origin. These aspects will be dealt with when they occur in the analysis of specific manuscripts. If it is established with a smaller or larger degree of certainty that a manuscript is not imported, but in fact locally made, one should examine if it is possible to see which European region the script and decorations are influenced from. The influence may stem from the scribe’s teacher, models in the surroundings, or the exemplar the scribe is copying from, elements from which would all be embedded in the copied product. In other words a manuscript displays “layers of influence” which may be complex and difficult to separate.

103

Part I: To study manuscript fragments Still, the features to look for are the same as those mentioned above; the shapes of the a and g, the ampersand and the punctus elevatus, and the other features. As concluded one may not get further than “this is not English or influenced from England”, but in the case of evaluating Norwegian script, this would be significant. The English influence on the Norwegian script has since long been accepted as a general truth, and although the English influence is certainly strong, it may not be the only one. Not only is the script in the earliest missals basically German, also further into the twelfth century, “French a’s” and “German g’s” occur in manuscripts presumed to be Norwegian (for instance Seqv 30 and Seqv 59b add from s. xii/xiii).

104

3.3. Grouping manuscript fragments

3.3. Grouping manuscript fragments The search for recognisable scribes or scribal features is currently largely dependent on luck, coincidence and visual memory, and is left more to chance than we would like. Still, that is the only alternative at the present stage of the research. An aim for this study as for future research is to sort fragments and manuscripts into groups, to make them more manageable as objects of study. The grouping of manuscript fragments can be done on different levels: 1) More than one fragment from the same manuscript. This was attempted when the fragments were first detached and placed in envelopes, but discoveries of this kind are still not unusual. The discussions around the possible connection between a fragment and an already registered “manuscript” also occur in this study (cf. Seqv 11). 2) More than one fragment with the same scribe from different manuscripts. The challenge is to exclude that the instances of “same scribe” is not merely an instance of “same manuscript” as in point 1). Several of the facsimiles in this study can work as illustrations to the few cases of undoubtedly Norwegian scribes identified in the Latin material, and presented by Lilli Gjerløw in her introduction to the edition of the Nidaros ordinal (Gjerløw 1968, 34-8). So far only a few hands have been identified as appearing in both vernacular and Latin written sources, but hopefully this number will grow. 3) The third level of manuscript grouping is the identification of a house style, or features in script and production so unusual that one may say with a fair amount of certainty that they probably originated from the same scriptorium. This, for example, is the method applied by Michael Gullick in his work in the Swedish National Archives (Gullick 2005). An interesting question regarding the “house-styles” is whether or not products are examples of different scribes writing in the same style, or the same scribe writing in a

105

Part I: To study manuscript fragments slightly different manner. An interesting case is that of the Icelandic Homily book. In this book the number of identified scribes has during the last thirty years been reduced from fourteen and twelve scribes (Rode 1974; Westlund 1974) to one single scribe in the latest edition (Leeuw van Weenen 1993). Kristin Bakken (1997) has discussed the difficulties of arguing for a scribe’s identity (in the sense if one document displays the same hand as another or not), and also treats the span of variety one should allow for in one single hand. The latest example of the reduction of multiple hands to one is in Michael Gullick’s forthcoming article on the Old Norwegian Homily book, where he argues that one scribe, not four, is responsible for the whole book, including a group of inserted leaves. Another question is whether we could expect scribes of very different styles writing in the same house. Parkes points to the example of the house of Cluny, a house which attracted scribes from all over Europe:, and where “As a result one manuscript may have been copied by a group of scribes whose hands exhibit wide regional differences;” (Parkes 1982, 23). Parkes describes scriptoria as historical phenomena, and emphasises the importance of considering the historical factors in the interpretation of the medieval books: Our ability to identify their products, to understand them, and even to date them satisfactorily, will depend on our appreciation of the historical factors involved. Different historical factors affecting different scriptoria produced different palaeographical features in their respective manuscripts. (Parkes 1982, 22).

A natural next step would be to link a group of manuscripts to a specific scriptorium or institution, like the ones listed in chapter 1.4.2. The problem, as Lapidge points out in his study of the Anglo-Saxon libraries, is that if the starting point is lacking, no linking chain can be constructed. The method “requires as a starting point that at least one manuscript can be reliably assignable to one centre, so that that manuscript can act as the first link in a subsequent chain of palaeographical evidence” (Lapidge 2006, 65). The hopes of finding a certain link to a scriptorium are even smaller when talking about manuscripts as fragments, when the chances of finding a colophon or

106

3.3. Grouping manuscript fragments an ex libris are minute. Still, the work should start by finding features linking one piece of parchment to another piece of parchment, increasing the information about both of them, whether they in the end can be assigned to a specific scriptorium or not. With the limited number of towns and institutions in medieval Norway, the task is not as hopeless as it would be for a larger and more populated region.

107

Part I: To study manuscript fragments

Summary and conclusion to Part I Since so few Latin manuscripts from medieval Norway are preserved, Norwegian manuscript studies are in practice fragment studies, at least as far as liturgical manuscripts are concerned. 150 years after the fragments were first discovered in the National Archives in Oslo, a large part of the 5-6000 single fragments from c. 1200 Latin manuscripts still remains unstudied. The greatest progress in fragment studies has so far been made in the fields of musicology and liturgy, particularly through the work of the liturgist Lilli Gjerløw. A new initiative from musicologists after the turn of the millenium resulted in a project to register the NRA fragments in a database. In connection with this project three international workshops were held, demonstrating the benefit of international and interdisciplinary collaboration. The international experts showed that it is possible to argue with some confidence not only which manuscripts were locally produced and which were imported, but to a certain degree also identify the place of origin of imported manuscripts. This way the small and detached fragments from medieval manuscripts can serve as physical witnesses to the development of Norwegian scribal culture as well as Norwegian contact with other European centres in the Middle Ages and mechanims of cultural transfer. The advances made in Latin palaeography through the NRA workshops are a most interesting development, particularly as it is now difficult to extract further information from other, secondary sources to Norwegian book culture, like inventories or legal documents. It is time to move towards the primary witnesses to the importation and book production in the Middle Ages for new information. The majority of fragments are from liturgical manuscripts, which can be seen as good representatives of the medieval book culture, since they were among the first books to be used and copied after the introduction of Christianity in Scandinavia, and outnumbered other book genres in the Middle Ages. One of the most important tasks in future research of medieval book and scribal culture is to achieve a better overview and increased knowledge of the surviving manuscript fragments.

108

Summary and conclusion to Part I Ironically, although less than 1% of the manuscript material which was in Norway at the time of the Reformation still remains, it has proven difficult to approach the material. For this study I selected the c. 70 manuscripts with sequences already identified among the large number of fragments. The sequence repertory of Nidaros has been extensively studied with regard to layers of influence from different European regions or centres. By selecting fragments with a liturgical genre which has received much scholarly attention, this study benefits from both former and recent musicological research. After a corpus is established, it is necessary to consider the date and origin of the fragments. Secondary literature is helpful with respect to dating, but much more elusive when it comes to determining an origin. The method applied in this study is based on the combination of six different elements, the shape of the a, the g, the ampersand, the punctus elevatus, the way the minims meet the baseline and the execution of the smaller initials. This is not a mechanical method and it should be used with caution. It is an attempt to form a tool in the process of developing an “eye” for identifying a scribe’s origin. In time it will hopefully be easier to relate scribes and fragments to specific historical contexts than it is at the current stage. Norway is a small country, and with merely five mainland bishops’ sees and thirty monasteries, it should be possible, as more fragments are studied, to link fragments or groups of fragments to historical institutions or environments. In the following section manuscript fragments are used as witnesses to contact with European centres. Although a limited number of fragments are presented in this study, it is an example of how knowledge of book and scribal culture can be increased through the analysis of liturgical fragments.

109

Part II: Analysis

Part II: Analysis This second part, the analysis, is the central part of this thesis, and is dependent on and supported by the introductory Part I on one side and the catalogue in Part III on the other. The most interesting aspect of the work with these manuscript fragments has been the process of trying to extract from fragments as much information as possible about the original manuscripts, their origin, use and historical context. Since the primary interest of this thesis is the early stages of book culture, c. 1100-1300, the fragments selected for analysis are all from the twelfth and thirteenth century. Of the seventy manuscripts with sequences presented in the catalogue, seven manuscripts from the twelfth century and twelve from the thirteenth have been selected for closer study, some Norwegian, others imported. At the outset of this thesis I asked how fragments from Latin manuscripts, particularly liturgical, could best be approached in a study of medieval book culture, and how the studies of such fragments could shed light on the cultural transfer between European centres and the northern periphery. In my opinion we are entering a new phase of the study of medieval book and scribal culture, and in the next few years we will have to balance a need for a better overview with a need for more close studies of fragments from specific manuscripts. We should work further to make the fragments searchable (in the continuation of the database project or other projects) and more easily available and identifiable (for instance through the publication of facsimiles), and at the same time not postpone the study of fragments from single manuscripts. An analysis should take into consideration both contents and palaeographical and codicological evidence, and attempt to connect the manuscript fragments to other known material. It is necessary to hold on to the international and interdisciplinary networks established during recent years. Piece by piece, fragment by fragment, in time it will be possible to get a clearer picture of the past and more specific knowledge about medieval book culture in Norway and the contact with European centres.

110

Although there was no clear hypothesis at the outset of this work, no expected results, it proved more difficult to relate manuscripts to each other in groups than I expected. This testifies to the diversity and extent of the early scribal culture, and the need for more studies in this field. While some of the manuscripts in the analysis are already known to scholars through the work of Lilli Gjerløw, in particular, and Espen Karlsen, others are studied here for the first time. The single chapters in this section, devoted to fragments from different manuscripts, are almost like separate articles. One analysis does not build on another, even though they are related through the same purpose and basic method. The structure of the chapters is not uniform – the analysis has been guided by the fragments and to what extent it has been possible to relate them to other manuscripts or historical circumstances. The conclusions drawn in Part II will be on two levels. One preliminary conclusion or hypothesis, which in some cases can be the starting point for a closer study, ends each chapter. Finally, the results of each chapter in the analysis will be the foundation for a more general conclusion regarding the information from these sources about books and scribes in the twelfth and thirteenth century.

111

Part II: Analysis

4. The twelfth century This study starts in the twelfth century, which one may say is to start in mid story, after it all began. It is clear from the remaining pre-1100 missals that when the first remaining leaves with sequences were written in the first half of the twelfth century, some kind of scriptoria had been active in Norway during the past half a century, if not longer, according to the results of the NRA workshops (Karlsen 2003). Unfortunately, no books or book fragments remain which can tell us which sequences were written, copied or sung in Norway before 1100, and no manuscript can provide the answer to which sequences might have been the first ones to arrive – either through the knowledge and voice of a cantor or between the covers of a book. Only a few years into the twelfth century Norway became a province of the Danish archbishopric of Lund, and whether remains of a missal from the second quarter of the twelfth century is in fact a physical evidence of Nidaros’ connection to Lund will be discussed below. Towards the middle of the twelfth century the future bishops and archbishops of Nidaros were being educated abroad. Archbishop Eystein was not the only Norwegian to receive his education in the Augustinian house of St. Victor in Paris. Three other Norwegians bishops are listed as canons of St. Victor, namely Eirik, bishop of Stavanger (1171-1188) and later archbishop of Nidaros (11891205), and Tore, archbishop of Nidaros (1206-1214), and another Tore, bishop of Hamar (1189-1196) (Johnsen 1975). The breviary-missal discussed in the first chapter below may well be the sort of book a travelling priest in the mid twelfth century would find useful and bring home with him. One twelfth century breviary-missal, by Lilli Gjerløw tentatively coined the “Utrecht breviary-missal” (Seqv 33a add), has been edited by Gjerløw (1979, 75-80) and will not be treated below. Still, it serves as a reminder that the Low Countries also may have contributed to the growing corpus of liturgical books in Scandinavia. Other books, to be discussed below, testify to English influence, which manifests itself not only through imported books, but also in the style of twelfth century Norwegian script and the sequences selected for the celebration of the church feasts. How does English influence relate to former and contemporary impulses from Continental Europe and a German sequence tradition?

112

4. The twelfth century Several of the following chapters will discuss the “power-struggle” between the English or Anglo-French and German impulses in the last half of the twelfth century as it manifests itself in particular manuscripts. Eleven manuscripts with sequences, ten transmitted as fragments and one as a codex (although rebound with later material), remain from the twelfth century. Most of the manuscripts (not counting the codex) are represented by only a few fragments each. The exception is Seqv 31 add (Mi 80), a missal from which fifteen fragments have survived, two of them whole leaves. The manuscripts below will be treated more or less chronologically in the following section. Table 5: Twelfth century manuscripts with sequences Catal. no.

Signature

Accounts for:

Genre

NF

NS

Origin

Date

Seqv 11

NRA, lat. fr. 497

Akershus 1614

Mi

2

1

England? (Norway?)

s. xii²

Seqv 16

NRA, lat. fr. 471 NRA, lat. fr. 251

Hadeland 1617

Unknown

2

2

England

s. xii²

Senja 1614, Nordland 1614

Br-mi

2

2

France

s. xii¹

Seqv 30

NRA, lat. fr. 236

Trondheim 1617, 1621

Seqv

2

5

Norway? (Sweden?)

s. xii²

Seqv 31 add, Mi 80

NRA, lat. fr. 235 a.o.

Trondheim 162130 a.o.

Mi

13+ 2

1

Denmark?

s. xii¹

Seqv 32 add

NRA, lat. fr. 261

No provenance listed

Seqv

2

2

France?

s. xii m.

Seqv 33a add, Br-M 5

NRA, lat. fr. 294

Hadeland 1619

Br-mi

4

1

The Low Countries?

s. xii m.

Seqv 37 add

NRA, lat. fr. 664

Stavanger len 1639

Seqv

2

2

France (or Norway?)

s. xii²

Seqv 40 add

NRA, lat. fr. 776 NRA, lat. fr. LR pk. 142

Søndhordland 1628 Bergenhus len

Seqv

2

1

England

s. xii²

Seqv

2

3

Norway

s. xii²

Man

cod.

1

Norway

s. xii²

Seqv 29, BrMi 2

Seqv 53 add Seqv 59b

Copenhagen, KB, NKS 32 8°

NF = Number of fragments NS = Number of sequences

113

Part II: Analysis

4.1. A sequence for mass and magic: Seqv 29 (Br-mi 2), s. xii¹ One of the oldest manuscripts with sequences is a breviary-missal (Oslo, NRA Lat. fragm. 251) in Lilli Gjerløw’s catalogue labelled both Seqv 29 and Br-Mi 2, from which only two leaves (17,5 x 11 cm) remain. Added on an initially blank page of this breviary-missal are two sequences, one of which has played a special role in Norway and Iceland as a protection against evil, namely Alma chorus domini (AH 53, no. 87). The fragments from this breviary-missal have been edited by Lilli Gjerløw (1979, 7475).

4.1.1 A twelfth century breviary-missal Two dates have been suggested for this manuscript, one in the first half, the other in the third quarter of the twelfth century. On the envelope in the NRA is written in old-fashioned orthography: “Beg. af 12. Aarh. Antag. af nordisk oprindelse” (“Beginning of the twelfth century, probably of Nordic origin”). The date seems to be correct, but the assignment of a Nordic origin is probably not. In Lilli Gjerløw’s edition she proposes the date 1150-75, and the notation identified as being of Norman origin, with reference to Georg Reiss (Gjerløw 1979, 74). The third quarter of the twelfth century seems a little late for neumes in campo aperto, which one would rather expect to find in the first half of the century. There are three text scribes, and three different types of notation present on the two leaves. The first scribe (i.e. the one presumably responsible for the main bulk of the breviary-missal) writes a notation looking almost like petits carrés in campo aperto, as if he would have used staves if there had only been room for them. In that sense Seqv 29 might be one of the many manuscripts from the first half of the twelfth century showing traits from the transition from neumes in campo aperto to musical notation on staves, which were at that time being introduced to give a more precise rendering of the melodic movement. When text scribes following a model with neumes did not leave enough room for staves to be drawn in, this posed a serious challenge to the music scribe.122

122

I am grateful to Susan Rankin for making me aware of this phenomenon in the transition from one system of music notation to another.

114

4.1. A sequence for mass and magic: Seqv 29 (Br-mi 2), s. xii¹ On the first leaf are parts of the liturgy for the octave of the Epiphany (6-13 Jan). 123

According to Gjerløw, the gospel antiphon Ordines angelorum (cao 4189)

for the

vespers appears in the same position in French and English uses, like SaintWandrille, Bec and St. Alban’s (Gjerløw 1979, 75). On the second leaf the postcommunio prayer Celesti lumine quesumus domine represents the last entry of the first scribe.

Fig. 1: Seqv 29 (Br-Mi 2), NRA Lat. fragm. 251-2r. The first three lines show the main scribe of the breviary-missal. A second hand has added text and music below. A third scribe entered the sequences on the verso-side of this leaf, see fig. 2 and the reproduction in Part III: Catalogue. 123

cao: Corpus Antiphonalium Officii (Hesbert and Prévost 1963)

115

Part II: Analysis

The second scribe added three responsories for the Holy Thursday of Easter week, the first one with neumes (Ecce vidimus eum, cao 6618; Una hora non potuistis, cao 7807 and O Juda qui dereliquisti, cao 7272). On the verso side of the same leaf the third scribe has entered two sequences for Pentecost: Sancti spiritus assit (AH 53, no. 70) and Alma chorus domini (AH 53, no. 87). Only the first three verses of Alma chorus fitted into the page, and the rest of the sequence was presumably completed on a now lost leaf. The scribe responsible for the added sequences (presuming in this case that the text scribe was also the music scribe) seems to have been writing on staves, which are practically invisible for most of the page. His type of notation, without the staves, is very similar to that in an eleventh century French gradual (Montpellier, Bibliothéque de l’École de Médecine, H. 159) presented in Paléographie musicale 8 (Mocquereau 1972, orig 1901). How can we know that the origin of the breviary-missal is France? All three text scribes share the same basic features: there is a sharp angular turn upwards to the right when the minim meets the baseline (more pronounced in the second and third scribe than in the first).124 This is rarely as pronounced in English or German manuscripts. The third scribe seems to write a slightly more old-fashioned hand than the first scribe, although they are contemporary. The shape of the ampersand, rather tilted to the right, finishing with a downward stroke rather than an upward one (

)

would also point in the direction of France, or possibly Germany. This would at least indicate that the scribes were not English, since English scribes would write a more upright ampersand, with the final stroke going upwards (

).125 As the contents point

to France or England, and the scribes are not English, it is very likely that the origin of the manuscript is France, and that the breviary-missal was written by a French scribe, with additions made by two contemporary scribes, also French. As the musical notation seems to be in a state of transition between neumes and stave 124

See for instance Derolez 2003, Pl. 6, and Thomson 1969, pl. 4 and 6. For the different shapes of ampersands, see Derolez 2003, 66-67. Although Derolez does not mention France specifically, he shows an example of the ampersand above in a French 12th century ms (ibid., Pl. 3). 125

116

4.1. A sequence for mass and magic: Seqv 29 (Br-mi 2), s. xii¹ notations, it is reasonable to suggest a date of origin in the first half of the twelfth century.

4.1.2. From France to Norway We cannot know for certain when Seqv 29 was brought from France to Norway. One argument that the manuscript was brought to Norway during the twelfth century is the notation, which for the main part of the breviary-missal was probably neumes in campo aperto. Only a few decades later one would expect notation to be on staves, and neumes would be old-fashioned. A Norwegian travelling in France, say, one century later, in the thirteenth century, would presumably not see this breviary-missal as a book worth spending his money on and bringing home with him. In the early to mid twelfth century, however, this is the very kind of book which would be desirable for a man with an ecclesiastical future to acquire when abroad. Not only is the octavo format very handy, the surviving parchment leaves are paper thin, i.e. the original manuscript would not have been particularly heavy. The book genre is also good for a “travel-book”, as the breviary-missal would contain all the necessary elements he would need for the celebration of both mass and office. Where was the book used in the Middle Ages? The two leaves were used to bind accounts from Senja 1614 and Nordland 1614, (“Nordlandene” comprised Salten, Senja, Andenes and Troms, cf. map in ch. 1.1.4.). The fief of Nordland was in 16041618 held by Hartvig Bilde, who at the same time also held Helgeland and Lofoten. It is likely that the manuscript was dismantled and used for bindings in northern Norway, although with only two leaves from the same year, Akerhus cannot be excluded as the place where the manuscript was reused. If the secondary provenance was in fact northern Norway, it is possible to speculate, as northern Norway in the Middle Ages was under the see of Nidaros proper, that the book was at some point brought north from the town of Nidaros.

117

Part II: Analysis 4.1.3. Alma chorus in Norway Both Sancti spiritus and Alma chorus domini were widely known sequences, and they were both later entered in the official Nidaros liturgy as represented in the ordinal. While Sancti spiritus is found in several Norwegian manuscripts, Alma chorus is only transmitted in this single manuscript. However, there are other examples of its use. Alma chorus, with its particular textual elaborations on the name of God, seems to have retained a special role as protector, both in regular sung form or through amulets and magical formulae (Gjerløw 1956, KLNM I, 92-94). Gjerløw points to an example of the force ascribed to Alma chorus as demonstrated in Sverre’s saga in the fourteenth century Icelandic manuscript Flateyjarbók (Copenhagen, Royal Library, GKS 1005 fol.) in connection with the seabattle in Bergen in 1181 (which is a rather interesting reference to a sequence in Western-Scandinavian literature, regardless of its historical validity). The account describes how king Sverre in the heat of battle puts down his weapons and on his knees and with his hands lifted to the sky sings “sequenciuna Alma chorus dei” [sic]126 from beginning to end without further protection (Unger and Vigfússon 1862, 583-4). Not only was Sverre not harmed, but his adversary, the young king Magnus Erlingson, was immediately after wounded in the foot, slipped on a pool of blood and fell backwards. And although Magnus insisted that he was not seriously injured, Sverre’s men called the victory.127

Fig. 2: Seqv 29 (Br-mi 2), NRA Lat. fragm. 251-2v, last lines (enlarged), showing the first three verses of the sequence Alma chorus domini.

126

The mistaken “dei” for “domini” in the text does not seem unnatural, given that the word “domini” was probably, as indeed in the case of Seqv 29, abbreviated to “dni”. 127 After fleeing to Stavanger, Magnus was advised by bishop Eirik to go back to Bergen and attack Sverre again, and although Sverre at that point fled Bergen, the struggle between Sverre and Magnus ended with Magnus’ fall in a battle a few years later, in 1184.

118

4.1. A sequence for mass and magic: Seqv 29 (Br-mi 2), s. xii¹ De Nominibus Domini 1. Alma chorus Domini nunc pangat nomina summi: 2a. Messias sother emanuel sabaoth adonai, 2b. Est unigenitus via vita manus homousion, 3a. Principium primogenitus sapientia virtus, 3b. Alpha caput finisque simul vocitatur et est omega, 4a. Fons et origo boni paraclitus ac mediator, 4b. Agnus ovis vitulus serpens aries leo vermis, 5a. Os verbum splendor sol gloria lux et imago, 5b. Panis flos vitis mons ianua petra lapisque, 6a. Angelus et sponsus pastorque propheta sacerdos, 6b. Athanatos kyrios theos pantocrator Iesus, 7. Salvificet nos, sit cui saecla per omnia doxa. On the names of God 1. May God’s blessed choir now sing the names of the highest one: 2a. Messiah, Sother, Emmanuel, Sabaoth, Adonai, 2b. He is the single son, the way, the life, the hand, the consubstantial 3a. The beginning, the firstborn, the wisdom, the virtue, 3b. the alpha, the head and the tail, also called omega, 4a. The source and the origin of goodness, consolator and mediator, 4b. The lamb, the sheep, the calf, the serpent, the ram, the lion, the worm, 5a. The mouth, the word, the splendor, the sun, the glory, the light and the image, 5b. The bread, the flower, the vine, the mountain, the gate, the rock and the stone 6a. The angel and the groom, and the shepherd, the prophet, the priest 6b. Athanatos, kyrios (Lord) theos (God) pantocrator (almighty) Jesus 7. May he save us, and may the glory be his for ever and ever.

Alma chorus domini also appears on three medieval amulets, which are treated by Lilli Gjerløw (1954) and which will only be briefly mentioned here. One is a lead cross from the island of Bru outside Stavanger (now in Stavanger, Arkeologisk museum), with the words “serpens aries leo vermis” from verse 7 incised in runes (Olsen 1911; 1954, no. 262-3). In addition a runic inscription containing words from verse 2 (“messias soter emmanuel sabaoth adonay”) is on a wooden amulet (now in Oslo, Museum of Cultural History, University of Oslo) discovered under the floorboards in 119

Part II: Analysis Borgund stave church in Sogn (Bang 1901, no. 1069; Olsen 1957, no. 348). Finally, Alma chorus appears on a parchment amulet from the fifteenth century (in Arendal, Aust-agder kulturhistoriske senter) containing the prologue of John (In principio) and other liturgical texts.

128

Fig. 3: Parchment amulet. Alma chorus covers the last part of the leaf. Arendal, Aust-agder kulturhistoriske senter, AA 186, skinnbrev, edited as DN 7, 441. Photo: Arendal, Aust-agder kulturhistoriske senter.

Lilli Gjerløw connects this magical use of Alma chorus with that of two other liturgical texts, namely the antiphon Deus pater piissime and the sequence Christe salvator. All three texts are presumed to be of French origin, and are found on lead crosses in burial sites in the Stavanger area. Gjerløw suggests that their presence in Norway may be the result of a link between northern France and Norway, and that they have been brought to Nidaros, possibly via Stavanger, by a Norwegian priest. The possibilities of this would be ample in the twelfth century, at a time when

128

The text is edited as DN 7, 441. It is also registered in Bang 1901 as no. 1071. Lilli Gjerløw has treated other items in Arendal, Aust-Agder kulturhistoriske senter, in an early article (Gjerløw 1959).

120

4.1. A sequence for mass and magic: Seqv 29 (Br-mi 2), s. xii¹ Norwegians looked to France for a higher education. Gjerløw points out that a man like Eirik Ivarssøn, bishop of Stavanger (1171-1188) and later archbishop of Nidaros (1189-1205), could be a candidate for the introduction of Alma chorus, Deus pater piissime and Christe salvator (Gjerløw 1954, 106). Since Alma chorus was made part of the official Nidaros liturgy, and is present in our twelfth century breviary-missal, it is likely that the sources for the pieces of text on the amulets are indeed, as suggested by Lilli Gjerløw for the case of the antiphone Deus pater piissime, the priesthood and their liturgical books (Gjerløw 1954, 85). In other words we may assume that Alma chorus initially reached Norway as a regular part of the liturgy, not as a magic formula. Although Alma chorus was known in most parts of western Europe by 1200, particularly France and England, in the North it has only been found in Norway and Iceland. Alma chorus is not registered by C. A. Moberg (1927) or during the recent registration of the fragments in the Swedish National Archives in Stockholm (Björkvall 2006, 58-9). As recorded by Bannister and Blume Alma chorus ceases to appear in German manuscripts in the twelfth century, and before that does not seem to have a strong position in the German sequence tradition (Blume and Bannister 1911, 154). Since we should always be careful to argue something on the basis of negative evidence, I will leave it at that, and simply conclude that Seqv 29 and the other remaining evidence of Alma chorus are valuable pieces of evidence for Norwegian contact with France during the twelfth century. And indeed, if Alma chorus domini was brought to Norway from France by the archbishops Eystein or Eirik, it is more than just a little ironic that the sequence should be given such a role in the saga of king Sverre, protecting him in the name of God. Sverre was the cause of both Eystein’s exile in England (1180-83) and Eirik’s exile in Denmark (1190-1202), and subsequently brought about the downfall of king Magnus, who since the time of his crowning by archbishop Eystein in Bergen in 1163 had embodied the archbishops’ hope of church and kingship in harmony.

121

Part II: Analysis

4.2. A grand-scale missal for a grand-scale church: Seqv 31 (Mi 80), s. xii¹ The second manuscript from the first half of the twelfth century is in many ways a contrast to the breviary-missal just presented. While Seqv 29 was a handy “travelbook” of French origin, Seqv 31 add is an unusually large missal (41 x 25 cm), probably written in Scandinavia. The missal is in Lilli Gjerløw’s catalogue registered as Mi 80, and is entered in the NRA database as Codex 61.129 In addition to the thirteen fragments kept in NRA (Lat. fragm. 235, new numbers Fr.512-524), one leaf is kept in the Danish National Archives (DRA Fr 3630-31, LR Kristianopel 16), and another in the University library of Trondheim (Gunnerus’ Library, fragm. 9).

4.2.1. English scribe, German neumes and punctuation At first sight the missal looks English. The scribe, writing in the second quarter of the 130

twelfth century (c. 1120-30)

writes an English-type script, and the initials, according

to Michael Gullick, are English in style. A closer look at the notation, on the other hand, complicates the picture: One scribe, writing notation on staves, writes AngloFrench notation, but on one fragment someone has entered German neumes in campo aperto, very well executed, probably to supply notation missed by the music scribe.

131

A feature in the punctuation also points to German influence. On the leaf in Trondheim the scribe uses an unusual punctuation mark ( ), resembling a punctus flexus (“seven and point”), but as this sign seems to be used instead of a punctus elevatus (“tick and point”) and not in combination with it, it would seem that it is not in fact a punctus flexus but an unusual rendering of a German type punctus elevatus, found in the model for that particular text.

132

According to Parkes this shape of the

punctus elevatus is typical for southern Germany (Parkes 1993, 281).133 In the NRA fragments the scribe uses the “normal” Anglo-French punctus elevatus: .

129

I thank Espen Karlsen and Gunnar I. Pettersen for the use of the unfinished NRA database. A date suggested by Michael Gullick and Teresa Webber. 131 I thank Susan Rankin for these observations. 132 For the punctus flexus, see Ker 1960a, 58-59, and Parkes 1993, 306. 133 A leaf from Munich Staatsbibliothek, cod. lat. monach. 12620, probably written in Ranshofen, Austria, has the same type of punctus elevatus (Thompson 1969, Pl. 33) 130

122

4.2. A grand-scale missal for a grand-scale church: Seqv 31 (Mi 80), s. xii¹

Fig. 4: Seqv 31 (Mi 80), Trondheim, Gunnerus’ Library, fragm. 9: Notice the punctuation marks shaped almost like a narrow "punctus flexus", for instance in line 2a or line 7a.

123

Part II: Analysis 4.2.2. A product of Lund? Where was Seqv 31 add written, and for whom? This could simply have been a case of an English scribe copying from a German exemplar, were it not for the German 134

neumes.

This mixture of influences detectable in the manuscript seems to echo

missals of the late eleventh century studied at the NRA workshops, where German script met English notation. The eleventh century missals of mixed influence have a secondary provenance in the Oslo area, where it is also likely that German features would be stronger than in other parts of Norway for geographical reasons. Seqv 31 add could possibly be a product of a Norwegian scriptorium in the Oslo area. Another, and perhaps more likely, possibility is that the origin of the manuscript could be connected to Denmark, because it is unlike anything previously seen among the Norwegian missals presumed to be locally produced.135 In the first half of the twelfth century Nidaros was a province of Lund in Denmark, where it is just as likely that English or English-trained scribes could have worked with Germans, and encounter both English and German models to copy from. This marvellous and large missal was not written to be used in a common parish church. A missal in folio format would be most suitable for a cathedral, and would at the time it was written not have many manuscripts which could compete with it. Whether or not the missal was produced in Lund, it was most likely used in Nidaros. That all fragments from this manuscript have been used in bindings from the Trondheim area (including the fragment in Copenhagen), and that one leaf still remains in Trondheim, makes a secondary provenance in Nidaros relatively certain. The link between Nidaros and Lund must have been fairly stable at that time, and a missal could for instance have been commissioned or purchased there by a bishop of Nidaros while visiting the arch see.

134

Although as pointed out by Michael Gullick, an English scribe of such quality writing in England would presumably not unconsciously copy the punctuation form. 135 First suggested by Michael Gullick at the NRA workshop in August 2006.

124

4.2. A grand-scale missal for a grand-scale church: Seqv 31 (Mi 80), s. xii¹ It is worth taking a closer look at some of the remaining manuscripts from twelfth century Lund. Medeltidshandskrift 5 in Lund University Library is a lectionary with a presumed origin in Lund in the second third of the twelfth century.136

Fig. 5: Lund University library, MS 5, f. 150v. Notice the punctuation mark, and the otherwise mixed influence in the hand, such as the English-looking ampersand and the rather straightbacked g Notice also the horizontal hyphens on the baseline. The picture, and pictures of the rest of the book, is available on-line at http://laurentius.lub.lu.se. 136

For a description of the manuscript see St. Laurentius online digital library: http://laurentius.lub.lu.se/volumes/Mh_5/

125

Part II: Analysis Immediately striking is the punctuation mark, shaped almost like the one found in the missal (cf. fig. 5, l. 12). Another feature the Lund manuscript and the missal have in common is the small, horizontal hyphens placed on the baseline. In general, the scribe of the Lund lectionary also seems influenced both by English and German features. His ampersand looks rather “English” in shape, and initial a’s are tall (cf. Ker 1960a), while the straight-backed g would not be found in English manuscripts, but rather connect the scribe to the German-speaking parts of Europe or parts of France. So far the contents of Seqv 31 add have not been considered. The one surviving sequence from the manuscript may bring us closer to a conclusion.

4.2.3. A German sequence The sequence on the leaf in Trondheim, Omnes sancti seraphim (for all saints, AH 53, no. 112), which accounts for the missal being listed as a manuscript with sequences, is not part of the Anglo-French sequence repertory, and may, like the neumes and punctuation, point to a German exemplar. The sequence is in Analecta hymnica ascribed to Notker Balbulus, and it is present in a number of manuscripts from Sankt Gallen. It has not been found in any French manuscripts (although in a printed missal from Rouen), and only occurs in one English manuscript – one of the Winchester tropers (Oxford, Bodleian Lib., Bodley 775 of s. xi) (Blume and Bannister 1911, 197-98). What is most interesting is that Omnes sancti seraphim is also present in the witnesses to the Liber ordinarius for the rite of Hirsau in southern Germany (Kruckenberg 1999, 189). The abbey of Hirsau was reformed in the late 1070’s, and this reform-movement had a considerable impact beyond the borders of the abbey. In her edition of the Nidaros ordinal Lilli Gjerløw pointed out that in addition to the influence from North-German rites in the Lund liturgy identified by Helander (1957, 252), Lund (and later Nidaros) was highly influenced by the liturgy of Hirsau (Gjerløw 1968, 87). The influence of the Hirsau liturgy on the Nidaros ordinal, most likely via Lund, is

126

4.2. A grand-scale missal for a grand-scale church: Seqv 31 (Mi 80), s. xii¹ traceable also in the sequence repertory of Nidaros (Kruckenberg 2006a, 32). When or how the Hirsau ordinal reached Lund and left its mark on Danish liturgy, or when it was introduced in Nidaros, is not known. Omnes sancti seraphim was not made part of the later Nidaros ordinal, and it is not represented in any other manuscript from Norway. Since it is not unlikely that Seqv 31 add was brought from Lund to Nidaros, it could possibly represent a physical testimony for a Hirsau-Lund influence on Nidaros by the first half of the twelfth century. With this in mind, Seqv 31 add should be subject to further study regarding its contents and its relationship with Hirsau. For now it is merely worth pointing out that the Hirsau-Lund connection could be a possible explanation for the presence of a southern German punctuation mark in a Scandinavian manuscript.

4.2.4. Further research To get further in the question of the Lund-Nidaros relationship it would certainly be useful to look for further evidence of books possibly produced in Lund in the NRA fragment-collection among the early twelfth century material. The features to be looked for would include script with English features but with traits like the “punctus flexus-like” punctus elevatus or the small horizontal hyphens on the base-line, and the use of j for i in medial and final positions, which is a feature in the Necrologium Lundense from Lund Cathedral (Lund University Library, Medeltidshandskrift 6) (Gullick 2005, 66). As it happens, this feature is also present in the Lund Lectionarium, in the final i in the word omnj (see fig. 5, l. 11). With the exception of i longa used after i (ij), this is not a feature common in the work of English scribes (Gullick 2005, 59). A few other items in NRA’s fragment-collection are worth mentioning in this context. There are two fragments (both 12 x 8 cm) from a large homiliary, originally in folio format (NRA, Lat. fragm. 28), containing parts from a sermon of Pope Leo (Sermo LXX. De passione domini). The manuscript was written in the first half to mid twelfth century in a basically English-type script, but with examples of i longa after m and n

127

Part II: Analysis in final position (mundanj, dilectissimj). Another missal-fragment (Mi 56, NRA, Lat. 137

fragm. 296 – old number) has i longa after t (multj).

Lilli Gjerløw also investigated

the possibility of liturgical books being brought from Lund to Nidaros. She suggested a Lund origin for two twelfth century antiphoner fragments, although of a different type than the ones discussed here (see Gjerløw 1979, 82, 38, 44-45, Pl. 12). A search for and investigation into the traces left by Lund in the remaining fragments of Norwegian books would bring new knowledge regarding internal Scandinavian networks, and the influence an arch see like Lund had on its provinces in the twelfth century.

137

I am grateful to Michael Gullick for making me aware of this fragment.

128

4.3. Two sequentiaries of different origin: Seqv 40 and 53 add, s. xii med.

4.3. Two sequentiaries of different origin: Seqv 40 and 53 add, s. xii med. In this chapter a closer look will be taken at the type of book reserved for sequences only, the sequentiary. While the manuscripts in the two last chapters may be physical evidence of a twelfth century contact with France and Denmark (Lund), these two manuscripts point further west, to the British Isles. The two twelfth century octavoformat sequentiaries are of a similar type. Seqv 40 add (NRA, Lat. fragm. 776) comprises two fragments forming ¾ of a leaf (c. 14,5 x 10,5 cm), while Seqv 53 add (NRA, Lat. fragm. Sandaakers kat. LR pk 142) has two cropped leaves (12,5* x 11 cm).138

4.3.1. An English sequentiary I have labelled Seqv 40 add a sequentiary even though there is only one sequence preserved, and no way to confirm that the chants preceding or following the sequence Nunc luce alma were sequences and not other types of chants. I have done this because the format and lay-out, in combination with a relatively large script, is very typical for small sequentiaries and would be unusual in a gradual, missal or breviary-missal. A missal in such a small scale, with such a large script, would simply have been too thick and bulky. The scribe of Seqv 40 is English, as shown by the shape of the g and the ampersand along with the general regular and “tidy” execution of the script. The smaller initials are pen-drawn, also an English feature, in two colours: red and yellow (or light ochre). The yellow pigment is fairly common in the twelfth century, but it is rarely seen later. The parchment is quite shiny and the flesh-side is clearly different from the hair-side, and it does not have the “nappy” quality on both sides, which, according to Michael Gullick, may be a characteristic for Scandinavian parchment (Gullick 2005, 59). In the case of Seqv 40 add the parchment indicates that this is not the work of an English scribe working in Norway, but a manuscript which was

138

* = cropped.

129

Part II: Analysis probably written in England in the third quarter of the twelfth century, and which someone later brought to Norway. The content of Seqv 40 is the sequence Nunc luce alma (AH 37, no. 276) probably for the feast Petri ad vincula (1 Aug). This sequence is according to Blume and Dreves often found in French and English sources, with no mention of German sources (Blume and Dreves 1961, 241). Form and contents alike indicate that Seqv 40 add probably belonged to the Anglo-French tradition. Nunc luce alma later became part of the Nidaros ordinal.

Fig. 6: Seqv 40 add, NRA, Lat. fragm. 776. An English sequentiary in Norway. Notice the pen-drawn smaller initials in red and yellow (or a light ochre)

130

4.3. Two sequentiaries of different origin: Seqv 40 and 53 add, s. xii med. 4.3.2. A Norwegian sequentiary The second manuscript, Seqv 53 add, is a sequentiary of the same format as Seqv 40, and modelled on books similar to it. The writing space of the two books is very similar: 10,3 x 7,3, cm for Seqv 40 add and 10,5* x 7,5 for Seqv 53 add (since one line on the top of the leaf, c. 1,5 cm, is cut from the writing space, the original writingspace would have been c. 12 x 7,5 cm). The scribe wrote an English style script, notice, for instance, the clearly eight-shaped g and the English-looking ampersand. The scribe writes relatively evenly, but there is something odd in the way the minims 139

meet the baseline, and particularly how serifs are added to the final stroke of the a.

The splits on some of the ascenders do not look particularly English, and the presence of an æ-ligature in verse 17 of the sequence Celica resonant (“Preclara qua140 lux veræ [sic, common form: vera] micat”) would be a very odd feature in an English manuscript of the late twelfth century.141 In Latin script the æ-ligature had at this point long since been replaced with the e caudata, which was also used in this manuscript. The presence of an æ-ligature indicates that this scribe was familiar with the Old Norse vernacular, where this ligature was still in use. The use of e caudata indicates a fairly early date for this manuscript, at least before 1200, but presumably earlier as the use seems consistent. The e caudata was no longer used by English scribes from c. 1170-80 (Ker 1960a). The smaller initials in Seqv 53 add are red, and although some appear to be pen drawn, others are simply pen-written (see plate in Part III: Catalogue). The one surviving initial is a simple red N, decorated with a disc (the grey is due to oxidation of the red pigment). Another feature worth pointing out regarding Seqv 53 add is the surviving sewing holes from the original binding (not visible below. For illustration, see Part III: 139

The model for the applications of the serifs seems to have been something like Pl. 86 in Thompson 1969. 140 Qua corrected from quia. 141 The æ is also used in the Old English vernacular, but somehow it seems that English scribes did not bring scribal characteristics from the vernacular into the Latin, at least not as frequently as it could seem that Norwegian scribes did. There is, however, an example of the Latin word benedicere spelt benædicere in the socalled Taunton fragment, four leaves from an Anglo-Saxon homiliary from the second half of the eleventh century (Taunton, Somerset County Record Office, DD/SAS C/1193/77) (Gretch 2004). I am grateful to Aidan Conti for discussing this matter with me, and bringing to my attention this example of Old English æ in a Latin word. The use of an æ in a manuscript should not be used on its own as an indicator of Scandinavian origin, but may still be used if accompanied by other evidence.

131

Part II: Analysis Catalogue). These show that the threads went through holes rather than the horizontal slits that seems to be more typical for Norway (at least during the thirteenth century). Since English books were normally prepared for binding with holes, this could also mean that not only the general features of the script, but also other aspects of the production of the book were influenced by England.

Fig. 7: Seqv 53 add, NRA. Lat. fragm. Sandaakers kat. LR pk 142-1. Sequentiary most likely written in Norway.

132

4.3. Two sequentiaries of different origin: Seqv 40 and 53 add, s. xii med. 4.3.3. A mixture of English and German sequences Apart from its obvious English influence and presumably Norwegian scribe, the contents of Seqv 53 add are particularly interesting. It contains three sequences for Christmas: Celica resonant (AH 7, no. 21/AH 53, no. 19), Natus ante secula (AH 53, no. 15) and Nato canunt omnia (AH 53, no. 24). What is remarkable is that Natus ante secula, which is the Notkerian Christmas sequence used in the German area, and Nato canunt omnia, which is the primary Anglo-French sequence sung for the first mass on Christmas morning, are found together in the same manuscript. Natus ante secula occurs in numerous German and some French manuscripts, but, according to Blume and Bannister, not in any English manuscripts (Blume and Bannister 1911, 20). Celica resonant, on the other hand, is not registered in any German manuscripts (ibid., 32). Although we cannot know exactly how the Christmas sequences were organized in the sequentiary, the Notkerian sequence Natus ante secula has the rubric alia, which means that it is an alternative sequence under the same rubric as Celica resonant, the Anglo-French sequence preceding it. Celica resonant and Natus ante secula may both be aliae under a rubric in nativitate domini with Nato canunt omnia as the first sequence, although this cannot be known for certain. What we can say is that Natus ante secula was placed in a secondary position in this sequentiary as an alternative sequence. Seqv 53 add could thus be an early manuscript witness to the phenomenon known from the later Nidaros ordinal: that Anglo-French sequences were favoured for the main feasts, and Notkerian or German sequences remained only in secondary position (Kruckenberg 2006a, 32-3). In the Nidaros ordinal Natus ante secula was never included among the Christmas sequences, and disappeared completely from the official Christmas celebration of Nidaros. Seqv 53 add, the only manuscript with Natus ante secula among the remaining Norwegian and Icelandic manuscripts, may, in other words, represent a transitional stage in the development of the Nidaros ordinal, when the German Christmas-sequence had been pushed back by a strong Anglo-French sequence tradition, but had not yet gone out of use.

133

Part II: Analysis

4.4. Norwegian sequentiary - or English missal? Seqv 11 (= Mi 133?), s. xii² This fourth section addresses one of the challenges of working with manuscripts in the form of fragments. How can one be sure if a fragment belongs to one particular manuscript or not? The fragment labelled Seqv 11 (NRA, Lat. fragm. 497, 1-2) at first sight seemed like a case of an English twelfth century scribe present in a very small unit of two fragments. The situation turned out to be more complex, and thus more interesting, after the NRA workshop in the spring of 2005. After comparing photocopies of Seqv 11 with those of one of the missal fragment treated in the workshop, Mi 133,142 it soon became clear that the scribe of Seqv 11 and that of Mi 133 (old numbers: Lat. fragm. 256, 354) was one and the same. This could open up two possibilities: The scribe in question could have written two different books; one sequentiary and one missal, or, Seqv 11 was simply a hitherto undiscovered part of Mi 133.

4.4.1. A scribe of unknown origin At the Oslo workshop of the spring 2005 the scribe of Mi 133 remained a mystery. Even though he displayed some typically English features, like the English kind of ampersand and the 8-shaped g, Dr. Teresa Webber expressed some doubt as to the English origin of the scribe. Michael Gullick also sensed a certain awkwardness in the initials and suggested that they might be a result of the scribe copying a smaller English book, causing the initials to look a bit out of proportion. According to Gullick the parchment of Mi 133 was not typical for England, and neither was the ruling, although Susan Rankin pointed out that the ruling was systematic and successfully organized for the staves of the musical notation. The English palaeographers agreed that the manuscript had been written in the third quarter of the twelfth century, but as for the origin of the scribe of Mi 133 they were not convinced that he was English, although heavily influenced by an English type script.

142

I am grateful for the use of the NRA database regarding Mi 133.

134

4.4. Norwegian sequentiary - or English missal? Seqv 11 (= Mi 133?), s. xii² Among the characteristic features of his hand making him recognisable as a scribe are the slight backward slope,

143

along with a slim, diagonal a, with its head only

slightly extending leftwards, and a very crisp and sharp or-ligature. There are at least three alternatives regarding the scribe of the Mi 133: 1) An English scribe working in England, a bit “off beat”. 2) An English scribe working in Norway, or possibly somewhere else, under different circumstances than he was used to, making the product appear not “typically English” (perhaps the parchment or the ink was slightly different, or the quality of the pen). 3) A Norwegian scribe having learned to write in an English manner, either in Norway or England.

4.4.2. One book or two? The conclusion we draw regarding the scribe actually depends to a certain degree on the conclusion we draw regarding the books. If one scribe is present in two different manuscripts in a relatively small collection like the one in the Norwegian National Archives, it may generally be considered as an indication that the scribe worked locally, since the chance of one scribe appearing in fragments from two different imported manuscripts is minimal. Not to say that such a scribe would necessarily have been of Norwegian origin and training - a scribe may very well have worked for a long time in Norway, but been trained elsewhere. One may presume that scribes of different origins worked in Norway over a smaller or larger period of time, which is why several other aspects of the evidence should be considered; initials or other decorations, the quality of inks and pigments, ruling, parchment and so on. If Seqv 11 and Mi 133 really are two different books, the chances are that they were produced in Norway, wherever the scribe was from originally.

143

The backward slope is a feature often seen in English manuscripts in the middle and late part of the twelfth century (Ker 1960a, 35-6).

135

Part II: Analysis Fig. 8: Seqv 11, NRA, Lat. fragm. 497. The remaning parts of the initial F in Fulgens preclara. Notice the wedgemotif in red, just to the left of the vertical blue line. Compare also the hand to that of Mi 133 below.

Fig. 9: Mi 133, NRA Lat. fragm. 3541 (new number: Fr.703). The difference in colour is mainly due to the reproduction being a colour xerox and not a photograph. Notice the general backward lean, the small head of the a and the crispness of the round r after o.

While the illustration from Mi 133 is also taken from a section with staff notation for the sake of comparison, Mi 133 is an interesting opportunity to see a scribe writing under staff notation as well as on regular lines (cf. fig. 11). The letters, apart from the difference in size are perhaps slightly narrower, and there is a higher occurrence of tall d, but other than that the characteristics are the same. What are the arguments in favour of Seqv 11 and Mi 133 being two different books? Two things are worth taking a closer look at, namely the format and the initials. The fragments from Mi 133 seem to be from a book of slightly smaller format than those of Seqv 11. The lines of Mi 133 are 1,5 cm high, while those in Seqv 11 are 1,6 cm. The columns also seem a bit smaller in Mi 133, although this could be simply variation between inner and outer column. As mentioned, the initials in Mi 133 have been said to look slightly out of proportion, as if they were copied from a smaller model. The initial in Seqv 11, a blue F with red decoration, unfortunately lacks the left part, and

136

4.4. Norwegian sequentiary - or English missal? Seqv 11 (= Mi 133?), s. xii² it is not easy to see to what extent it would have looked compared to Mi 133 when whole. The remaning part of it, though, shows a particular feature, namely the wedge-motif (see the red wedges in fig. 8). This motif is found particularly in northern English manuscripts (Gullick 1998, 248). None of the initials in Mi 133 contains this motif. The person doing the initials could be another person than the text scribe, but not necessarily. The initial in Seqv 11, although it seems different from those in Mi 133, have some forms in common with this, like the curling, vegetal elements. It is likely that the initials were made by the same person, which could possibly be the scribe himself. Mi 133 contains another motif as well, namely small “lace-tongues”, single, double or triple (see fig. below). This motif is also found in books from St. Alban’s (cf. Thompson 2006, 65, fig. 46).144

4.4.3. Circumstantial evidence What are the arguments for Seqv 11 originally being a part of Mi 133? During a visit in the NRA in May 2006 it became clear that the envelope marked Lat. fragm. 497 did not have two fragments as expected, but six. While two of these were our Seqv 11, a third fragment clearly seemed to belong to Mi 133. The scribe was the same as the one in Seqv 11 and the other fragments from Mi 133, and the initials were of the same distinct kind as those in Mi 133. Three other fragments were of a similar kind, but done by a different scribe in a distinctly smaller format, so there is no reason to assume that the three last fragments were related to either Seqv 11 or Mi 133. All the fragments in the envelope marked 497 were used as bindings on accounts from Akershus 1613-1619. When comparing the regional connection of the tax accounts Seqv 11 and Mi 133 were used to bind, the list is as follows:

144

I am grateful to Michael Gullick for making me aware of this feature. It is also found in another fragment in the NRA, namely Lat. fragm. 672-1 (new number Fr.570), from an English missal from the second half of the twelfth century (Mi 94)

137

Part II: Analysis Mi 133: Lat. fragm. 354, 1-4: Hedmark 1616, Hadeland 1616, Gudbrandsdalen 1716 (1616?), Mossedal 1618 Lat. fragm. 356, 1-5: Follo 1612, 1613, Akershus 1613 Lat. fragm. 497, 3: Akershus 1613 Seqv 11: Lat. fragm. 497, 1-2: Akershus 1614 (For the geographical position of the different fiefs, see map in chapter 1.1.4.) The different place-names on the accounts of Mi 133, as listed above, were all part of the main fief of Akershus. Since Akershus is the common denominator for these accounts, it is likely that the bindings were done in the Castle of Akershus in Oslo. Consequently, it seems that Mi 133 suffered its fate in the Castle, where the leaves were removed from the manuscript and reused between 1612 and 1616 (disregarding the year 1716, which is probably an error). Seqv 11 was removed from an account marked Akershus 1614, which corresponds with the time and place when Mi 133 was dismantled. Although there was probably a large number of books at the Castle of Akershus for this purpose in the second decade of the seventeenth century, it becomes unlikely that we are talking about two different manuscripts. Regarding the difference in format and style of the initials it seems we have to find different explanations than two different manuscripts. As long as the variations in the height of the lines and width of the columns are measurable in millimeters, the difference may be due variations from quire to quire. It is likely that the two fragments of Seqv 11 was part of a sequentiary placed at the back of the missal Mi 133. The execution of the initials in Seqv 11 and Mi 133 is fairly similar, in spite of the different motifs used, and it seems likely that the same person made the initials in both Seqv 11 and Mi 133. If so, the initials of the main part of the missal and the initials in the sequence section could be influenced by their respective exemplars. The person responsible for making them (who may or may not have been the text scribe) may have used one model for the missal itself and another for the sequentiary part of

138

4.4. Norwegian sequentiary - or English missal? Seqv 11 (= Mi 133?), s. xii² the missal. While the missal may have been copied from a smaller English model, giving the initials an “out of proportions look”, the initials in the added sequentiary may have been inspired from a model of northern English origin.

Fig. 10: Mi 133 (NRA Lat. fragm. 354-1, new number: Fr.703). The size is reduced.

139

Part II: Analysis 4.4.4. Norway or England? Are there any indications to suggest where the manuscript was made? The first impression is that the colours are of good quality. The red seems a nice, clear vermilion, and the blue has a very nice hue – not at all like the greyish, greenish or uneven blues often seen in the Norwegian fragments. The third seagreen-looking colour (malachite green) is rather peculiar and not the commonly seen olive-green, but it seems even and of good quality. Looking at the pigments in the remaining initials of the medieval manuscript material in Norway, it seems clear that good quality pigment was not always available in many scriptoria. In the thirteenth century, and also later, initials are often only red, a colour any liturgical scribe would have available for the rubrics, or sometimes red and green rather than the expected red and blue. This could have been a question of price and availability. There were alternative recipes for red, but a stable blue colour seems to have been more difficult to obtain. The blue colour in Seqv 11 on the other hand, seems to be exactly of the same kind as that used in the contemporary English manuscript (of uncertain genre) labelled as Seqv 16. Here also, the blue is put on in a relatively thin layer (or is worn thin in a similar way), allowing the parchment to shine through.

Fig. 11: To the left: A segment from Seqv 11 (Lat. fragm. 497). To the right: A segment from Seqv 16 (Lat. fragm. 471), written in England in the second half of the twelfth century. The blue of the “I” in Seqv 11 is of the same thin, see-through character as the blue in the “M” to the right. The red also seems of a similar hue and quality. (There might be some variation in colour compared to the original fragments).

Although it is difficult to make a firm conclusion regarding Seqv 11, it does seem likely that the two fragments labelled Seqv 11 were once part of Mi 133, possibly as part of a sequentiary added to back of the missal. Looking at the pigments and the

140

4.4. Norwegian sequentiary - or English missal? Seqv 11 (= Mi 133?), s. xii² general good quality of the manuscript, I would hesitate to label this a Norwegian product, both in the sense that I do not think that the scribe is Norwegian, and I do not think that it is made by an Englishman working in Norway. I tend to believe that this was an Englishman working in England, perhaps in a smaller community somewhere, and that the book came to Norway at a later point. But wherever the manuscript was produced, Seqv 11/Mi 133, along with the mentioned Seqv 16, point to contacts with England.

141

Part II: Analysis

4.5. German sequences in Anglo-French script: Seqv 30, s. xii² The manuscript fragments labelled Seqv 30 (Oslo, NRA Lat. fragm. 236) are peculiar in more ways than one. First, there is the question of the script, which is unusual, and seems to be a mixture of English and French features. Secondly, the large format: Seqv 30 was a book in folio format with leaves 32,5 cm wide (height unknown, but probably more than 40 cm). Thirdly, there are the contents, and more specifically the contents in combination with this particular script: while the script is “AngloFrench”, the sequence-repertory seems to be that of the German-speaking areas of continental Europe.

4.5.1. German sequences for Easter and the Ascension To begin with the contents, the remaining sequences are as follows: Fr. 1r: Agni paschalis esu

(AH 53, no. 50)

(Easter-week)

Fr. 1v: Laudes Christo redempti

(AH 53, no. 45)

(Easter-week) (Easter-week)

Victime paschali laudes

[lacuna – the following sequence begins with a “round” letter, C, G, O or S] Fr. 2r: Rex omnipotens

(AH 53, no. 66)

(the Ascension)

Fr. 2v Summi triumphum

(AH 53, no. 67)

(the Ascension: alia or octava)

Several of these sequences are ascribed to Notker. The Notker sequences are also the core of the Hirsau repertory (cf. Kruckenberg 1999, 192). Laudes Christo redempti and Summi triumphum are both sequences which are not part of the “standard” AngloFrench repertory, but which did become part of the Nidaros ordinal. The Notkerian Easter-sequence Agni paschalis esu is also not part of the Anglo-French repertory, and is otherwise unknown in the Norwegian or Icelandic sources, including the Nidaros ordinal. As several of the sequences in Seqv 30 are Notkerian, it would be natural to look to Sankt Gallen. In the German tradition, the primary Easter sequence is Laudes salavatori voce, followed by a sequence for each day of the Easter-week. Since we do not know which days the sequences above belong to, it is difficult to enter Seqv 30 into a table. In the table below the first column shows the sequences for Easter week

142

4.5. German sequences in Anglo-French script: Seqv 30, s. xii² and the Ascension as they appear in the Nidaros ordinal (c. 1200), then follows the sequences of an eleventh century sequentiary from Sankt Gallen, Cod. 376 (which in contents is equal to another of the eleventh century Sankt Gallen manuscripts, Cod. 380). Finally, there is the Schaffhausen Ministerialbibliothek Ms. 95, a manuscript of the Hirsau tradition. Lori Kruckenberg’s comparison of the Hirsau sequentiaries shows a remarkable correspondence between the first four sequences of the Easterweek, she suggests that the organisation of the sequences during Easter may be seen as a “fingerprint” of the Hirsau sequentiary (Kruckenberg 1999, 201). In eight compared sources the first four sequences for Easter are Laudes salvatori, Pangamus creatoris, Agni paschalis and Grates salvatori (cf. Sch 95 below) (Kruckenberg 1999, 2023). Table 6: Sequences of Easter-week and the Ascension Easter-week and Ascension: Resurr. dni

Ordo Nidr. Eccl.

feria II feria III feria IV feria V feria VI feria VII Octava paschae

Zima vetus Prome casta Laudes Christo red. Dic nobis Sancta cunctis Pangamus creatoris Laudes salvatori

In Ascensione domini Dom. post asc.

Rex omnipotens

SG 376 (Sankt Gallen) 145 Laudes salvatori Pangamus creatoris Laudes redempti Christo [sic] Is qui prius Christe domine Agni paschalis Grates salvatori Laudes deo Carmen suo Hec est sancta sollemnitas Summi triumphum

Summi triumphum

O quam mira sunt

Fulgens preclara Victime paschali (vesper)

Sch 95 (Schaffhausen)146 Laudes salvatori

Pangamus creatoris Agni paschalis Grates salvatori Victimae paschali Laudes deo Carmen suo Laudes Christo red. Summi triumphum

If we keep in mind the order of the sequences in Seqv 30, Agni paschalis, Laudes Christo and Victime paschali, it is clear that it does not correspond completely with neither of the above. Seqv 30 does not follow the Nidaros ordinal, it is not copied from a Sankt Gallen manuscript, and does not seem to be following the sequence tradition of Hirsau, although the Schaffhausen manuscript contains most of the

145

The manuscript SG 376 is consulted on line, on the digital abbey library of Sankt Gallen, Codices Electronici Sangallenses (CESG): http://www.cesg.unifr.ch/en/index.htm 146 This example of a Hirsau sequentiary is taken from Kruckenberg 1999, 202. Sch 95: Schaffhausen Ministerialbibliothek, Ms. 95.

143

Part II: Analysis sequences in Seqv 30, with the exception of Rex omnipotens. It is possible that Hirsau has influenced the selection of sequences to a certain degree, but that Seqv 30 represents a local redaction. As it happens, all the sequences in this manuscript are also found in Swedish manuscripts (cf. Moberg 1927, I, 40-54; Björkvall 2006, 52-3).

4.5.2. “Anglo-French” script As the contents are “German”, the “Anglo-French” script in which Seqv 30 is written is all the more remarkable. The a is of a French type, which is not very common in the Norwegian fragments, except in the display-script of the first few words of a new sentence. The e is also unusual, as the mid-stroke is almost horizontal, making the e sometimes look like an uncial e, particularly in the cases when the lobe is not completely closed.

147

These features in such a large format, and in combination with a

round d make the pages look almost like they are written entirely in display-script. The g has the Anglo-French ductus and is left open, with the lower lobe turned downwards. While the a points to France, the capitals or litterae notabiliores are in a few cases pen-drawn, not pen-written, which is a feature of English books. The larger initials are surprisingly poor, done with what appears to be poor quality pigments. In addition to the brownish and orange red there is a turquois-looking colour, and something which could be described as ash-grey. While the script is unusual, but done by a relatively confident scribe, the initials seem to be of a considerably lower standard. This means that they could have been done by the scribe himself, who was not so used to that kind of work. It certainly makes it less likely that an expert did the initials. There is plenty of space between the lines of text, but not enough to put in staves. It is also very unlikely that anyone ever intended to put in neumes at such a late date. It could, I presume, be a very late example of someone copying a neumed model, without worrying about the notation. A final explanation is that it was never intended for music (which again would be rather unusual in the twelfth century). 147

There are some scattered examples of a similar type of unclosed, uncial-looking e. See for instance Gjerløw 1968, pl. 17.

144

4.5. German sequences in Anglo-French script: Seqv 30, s. xii²

Fig. 12: Seqv 30, NRA Lat. fragm. 236. Notice the initial S of Summi triumphum, which is not of particularly high quality. Notice also the French shape of the letter a and the horisontal stroke of the e, making the script resemble display-script.

4.5.3. A local product of mixed influence The manuscript from which the remaining fragments came was clearly intended to be grand and impressive. Any manuscript in a folio format, written in a large script, would be expensive, and it would probably be made in such a large scale so that more than one person could read from it at once. This, in turn, limits the number of options for which churches it was intended for. Seqv 30 was not written to be used in a small parish church, but in a larger church, possibly in one of the towns. The remaining parts of this manuscript were used to bind accounts from Trondheim for the years 1617 and 1621, and, although two fragments are not enough to make a firm

145

Part II: Analysis conclusion, it is likely that it was dismembered in Trondheim, and, consequently, that it was taken somewhere nearby. If the manuscript was in Trondheim or the Trondheim area in the early seventeenth century it is also likely that Nidaros was its medieval provenance and possibly its origin, as Nidaros was presumably an important centre of books production with several scriptoria in the late twelfth century. Could Seqv 30 be an early example of grand-scale books produced in Nidaros? The unusual script with a mixture of different elements, along with a discrepancy between the form and the expected contents, suggest at least that the origin was Scandinavian, and that the fragment reveal impulses from France, England and Germany.

146

4.6. A priest’s manual in German script: Seqv 59b add (ManNor F), s. xii²

4.6. A priest’s manual in German script: Seqv 59b add (ManNor F), s. xii² In 1962 the first volume of the liturgical books of Nidaros in the Series Libri liturgici provinciae Nidrosiensis medii aevi: Manuale Norvegicum (Presta handbók) was published (Fæhn 1962). The edition was based on three codices of Norwegian origin in the Royal Library in Copenhagen, Thott 110 8°, NKS 32 8° and NKS 133 f 4°. Of these the two last-mentioned contain sequences, and are therefore part of this study. The one treated in this chapter is the NKS 32 8° (Seqv 59b add), which is a composite volume with pieces from at least two books, the younger part of which is dated to before 1300. The codex was previously described by Kr. Kålund (1900, no. 924), E. Jørgensen (1926, 222) and Skånland (Fæhn 1962, xiii-xiv). The first two quires, including an inserted leaf (f. 5) are of a younger date than the main part, which also has lost a few leaves and a quire (Fæhn 1962, xiii).

4.6.1. Before 1200 The larger part of the manual (and the part containing the sequence) is dated by Seip to c. 1200, and he points to a range of suggested dates from the late twelfth to the early thirteenth century (Fæhn 1962, xxvii). I would be inclined to date this book to the second half, or last third or quarter of the twelfth century, because of the neumes in campo aperto, cupshaped or wavy suspension signs, and, most importantly, a regular use of e caudata.148 Although scribes abandoned these traits later on the European Continent than in England, one would expect them to be gone by 1200 (cf. Derolez 2003).149 There are no hints of Gothic traits in this script; no biting and no round r except after o. Even when allowing for delays due to a provincial origin, it seems likely that this manuscript was written before 1200. Before returning to the older main part of the book, it is worth remarking that the younger parts of the book are dated by Seip to before 1300 (Fæhn 1962, xxv). The scribe wrote a very uneven hand, which makes dating more difficult. As the 148

Seip points out that sometimes the e caudata has the tail above and to the right of the letter, for instance on f. 16r (Fæhn 1962, xxvii). 149 Although there is a German example of an e caudata in 1240 in Thomson 1969, Pl. 37.

147

Part II: Analysis ampersand is used for “et” and there are no signs of biting, but there is still a squareness to the script and breaking of the arches of m’s and n’s, I would think that the scribe was working around the second third or the third quarter of the thirteenth century. The scribe writing the fifth folium in the book looks older, and this leaf could possibly date to the first half of the thirteenth century; it is a nice and even hand, the d’s are straight, and there is no biting. The tironian note for “et” is uncrossed, and the punctuation is placed mid-line, not on the baseline. When these later leaves were bound together with the remains of the older manual is not certain, but they are presumed to have a secondary provenance in the Trondheim area because of a Norwegian marginal notice in the dialect of that part of Norway (Fæhn 1962, xxvi). The oldest part of NKS 32 8° is the focus of this chapter, as this is the part holding the sequence. The scribe is basically writing a German-type script, as shown by the straightbacked g, although he is writing rather unevenly. The pen-written smaller initials are touched with red, a feature of German manuscripts. Worth pointing out is also the “punctus flexus-like punctus elevatus” like the one in chapter 4.2. Here the situation is the same: Although it looks like a punctus flexus, it is used as a punctus elevatus, and is therefore another example of the German-type punctuation.

Fig. 13: Seqv 59b add, Copenhagen, Royal Library NKS 32 8°, f. 34r, top part of the page (for colourplate of the sequence-page, see Part III: Catalogue). See the special form of the punctus elevatus in line 1 and 3. Notice also the pen-written initials, touched with colour, which is a German or Continental feature. Picture taken from Fæhn 1962.

148

4.6. A priest’s manual in German script: Seqv 59b add (ManNor F), s. xii² 4.6.2. Germany, Denmark or Norway? Could this be another example of influence from Lund? It would be difficult to suggest this based on this punctuation alone. If this feature appeared in combination with English scribal features, or an unexpected i longa for instance, the case would be different, and one would have arguments to suggest such a link. In this case we do not. Was the book written in Germany or Norway? The general quality of the book would not indicate a larger scriptorium, but rather somewhere in the periphery, perhaps even that the book was written by the person needing it (the peculiar random notation could perhaps indicate that personal “need” was governing the selection). The notation is regarded by the musicologist Bruno Stäblein as of a German type (Fæhn 1962, xxxii). One element in the contents points to Norway, namely the prominent position of St. Olav in the litany. St. Olav is placed first among the saints, only after St. Stephanus the protomartyr (Fæhn 1962, 168). The litany could also hold the clue to which church the manual was written to be used in, since it prays for the “clerum et plebem sancte marie sanctique nicolai” (Fæhn 1962, 170). This unfortunately does not help us very much, since St. Mary and St. Nicholas are two very common saints for church dedication, although relatively few of the original dedications of churches are still known.

4.6.3. A Notker-sequence NKS 32 8° is entered among the sequence manuscripts because of the presence of one sequence, namely the Notkerian Easter-sequence Haec est sancta sollempnitas sollempnitatum (AH 53, no. 56), which is otherwise not found in the Norwegian or Icelandic material, and which was not included in the Nidaros ordinal. As shown in Table 6 in chapter 4.5., this sequence is in Sankt Gallen assigned to the octave of Easter, while here it appears to be used for Easter Sunday. Blume and Bannister list a large number of German manuscripts with this sequence, and a few non-German (Blume and Bannister 1911, 99-100). The text in our manuscript follows the edited text faithfully with the exception of verse 9, where iuge famulantur has become

149

Part II: Analysis famulantur iuge, but the scribe (presumably the same hand as the main text) has marked famulantur with a suprascript b, and iuge with a suprascript a, indicating that the words should switch places. It is curious that only the first parts of verses 8 and 9 are equipped with neumes. It is noteworthy that this is a manuscript clearly influenced by Germany at a time when the English and French influence seem to have taken a firm hold of both script and liturgy. It would be interesting to come closer to the circumstances for the making of this manuscript. The scribe could be a German writing in Norway, or a Norwegian trained in Germany. NKS 32 8° does contribute to the impression that the twelfth century was a complex one in respect to the influences on book culture, both regarding imported books and local production.

150

Conclusion for the twelfth century

Conclusion for the twelfth century It is remarkable how wide the variation is for the eleven manuscripts preserved from the twelfth century with respect to their genre, size, quality and influence. The expected link with France is testified through a breviary-missal in “travel-book” format, containing the Godsname-sequence Alma chorus domini which seems to have played a role not only in the official liturgy of the church, but also in the more popular beliefs regarding the protection from danger or evil. A grand-scale missal may be one of the few books showing a connection to Lund during the very time Nidaros was a province of Lund arch see. Two small sequentiaries may, each in their own way, show links to England, one through import and the other through English influence on local book production, and even a stage in the process of a German sequence-repertory being surplanted by an Anglo-French one. The German sequences and Anglo-French book-culture also seems to meet (and merge) in a largescale sequentiary produced in Scandinavia, possibly in Nidaros, at the end of the century. The books most likely to be brought from abroad are not all in a “handy” size. While the breviary-missal from France and the sequentiary from England could easily have been brought in someone’s luggage (or pocket), this is not the case for the two missals, one from the Low Countries and the other possibly from Lund. People acquiring books abroad would have bought books according to their wealth and rank and their type of luggage. A student and a bishop would be two different kinds of “tourists”, and the books in the student’s “backpack” would presumably not look like those found in the bishop’s book chest (which he of course did not need to carry himself). Among the books presumed to be locally produced there is also a considerable variety: The earliest sequentiary is small and practical, while the other is large and ambitious. One is in a “standard” English script, with an indication that its scribe also wrote the Old Norse vernacular, while the other is in an unusual Anglo-French

151

Part II: Analysis type script. That the influence at the end of the twelfth century was not only from the west is shown by the surviving manual, highly influenced by German script.

Map 3: The Nidaros archbishopric (marked in darker colour) and a selection of European centres in Scandinavia, England, France and Germany with a relevance for Nidaros.

152

Conclusion for the twelfth century Several different European centres or regions are represented in the Norwegian twelfth century book culture, either through direct contact and imported books. Others are present once or twice removed, such as the influence of Hirsau which can be traced in the ordinal of Nidaros in general, in the sequence repertory, and possibly even in peculiar southern German scribal characteristics, presumably channeled northwards through Lund. The manuscripts treated in the first part of this analysis certainly give the impression of Scandiavia as a melting pot: a place where different impulses from different places met and merged.

153

Part II: Analysis

5. The thirteenth century The thirteenth century is in many ways fundamentally different from the twelfth. The ecclesiastical institutions were now “old” and well-established, and during the beginning of the century, in the second decade at the latest, a general liturgical ordo for the whole Nidaros see was issued. What kind of activity would such an event produce? Was there large pressure to produce books in correspondence with the ordinal, or would people amend and correct the ones they had (or alternatively – go about business as usual)? What would be the effect on the import of books, which we have just seen was extensive in the twelfth century? The manuscripts with sequences in the twelfth century appeared to be imported from several different countries in western Europe, from France, England, the Low Countries and Denmark. At the same time the first examples of presumed Norwegian products showed a wide range of different scribal features – from England, France and Germany. In the thirteenth century a new situation emerges. Most of the investigated sources from this century appear to be local products, which was not expected. Not to say that there was no import of books in the thirteenth century – this only means that these imported books are not part of my group of manuscripts. I will argue in each case what indicates a Norwegian (or Scandinavian) origin. In the thirteenth century one may assume that the “learning-process” was over, and that Norwegian scriptoria, possibly including secular workshops, were up and running. One should think that when there are so many different examples of local book production from the same century, it should be possible to put some of the manuscripts into groups which have features in common. This has turned out to be difficult. In fact, it has been easier to contrast some of the material.

154

5. The thirteenth century Table 7: Thirteenth century manuscripts with sequences Catal.

Signature

Seqv 1

NRA, lat. fr. 418

Seqv 2

NRA, lat. fr. 336

Seqv 3 Seqv 6

NRA, lat. fr. 284 NRA, lat. fr. 911

Seqv 10

NRA, lat. fr. 427

Seqv 15

NRA, lat. fr. 419

Seqv 18, Mi 106

NRA, lat. fr. 881, NRA, lat. fr. 882, NRA, lat. fr. 932

Seqv 21 Seqv 22a, Man 19 Seqv 22b, Mi 112 Seqv 23, Gr 26 Seqv 25

Accounts for:

Genre

NF 150

NS

Origin

Date

Numedal 163033, Eiker and Brunla 1631-41 Hadeland 1621, Valdres 1628 Hadeland 1625 Trondheims len 1611 Numedal 1628

Seqv

46

29

Norway

s. xiii¹

Seqv

2

6

Norway

s. xiii¹

Seqv Seqv

2 2

3 3

Norway Norway

s. xiii¹ s. xiii m.

Unknown

2

2

Norway?

s. xiii

Fredrikstad 1615-17, Tunsberg 1613 Trondheims len 1625, 1626, Tr.heims gård 1624, Tr.heims len 1625, 1626

Seqv

7151

5

Norway

s. xiii¹

Mi

7+6+1

4+1152 +1

Norway

s. xiii m.

NRA, lat. fr. 715 NRA, lat. fr. 1007

Nordhordl. 1628 Andenes 1624, 1632, Nordland 1628, 1629

Seqv? Man

2 8153

2 1

Norway? Norway

s. xiii¹ s. xiii²

NRA, lat. fr. XIX

Bratsberg len 1639-1641

Mi

2

1

Norway

s. xiii²

NRA, lat. fr. 281

Akershus Slott 1604

Gr

2

1

Norway?

s. xiii¹

NRA, lat. fr. 532

Fredrikstad 1611

Seqv

2

2

s. xiii¹

Seqv 27 Seqv 33b add, Mi 107

NRA, lat. fr. 870 NRA, lat. fr. 489

Unknown Mi

2 17

1 5 (inc)

s. xii/xiii s. xiii¹

Seqv 38 add, Br-Mi 3

NRA, lat. fr. 668, S.kat: LR pk 288, S.kat: LR pk 34, SAS, fragm. no 13

Stjørdal 1632 Oslo and Hamar 1575, Gudbrandsdalen 1607, Tønsberg 1638 a.o. Stavanger len, Jæren og Dalene 1651-2, Ryfylke 1650, Kirkeregnskap for Stavanger len 1652-4

Norway or England? Norway? Norway?

Br-mi

2+2+2 +1

3

Norway

s. xiii m.

Seqv 39a add

NRA, lat. fr. 750

Seqv

2

3

Unkn.

s. xii/xiii

Nordfjord len 1614

150

Fragm. no 39 and 46 contain tropes, not sequences. 5 of 7 fragments have sequences. 152 Virgini Marie laudes appears as incipit. 153 From a total of 8 fragments, only one fragment has sequences. 151

155

Part II: Analysis NRA, lat. fr. 764

Bergenhus len 1612

Mi

2

1 (inc)

Norway

s. xiii¹

NRA, lat. fr. 798

Nordfjord 1644

Unknown

2

1

France (or Scand.?)

xiii¹

Seqv 45b add, Man 1 Seqv 46 add, Mi 49

NRA, lat. fr. 1028

Nordlands len 1628, Senjen 1629-33

Mi/man

6

1

Norway

s. xiii²

NRA, lat. fr. 1029

Nordlands len 1614

Mi

2

1

Norway?

s. xiii¹

Seqv 51 add

NRA S. kat: LR pk 692

B. Knekteskatt 1611-13 Bamble

Unknown

1 fol.

1

Norway

s. xiii¹

Seqv 54 add Seqv 55 add Seqv 57 add

Arendal, AA72

Tags from Bygland

Unknown

2 tags

3

Unkn.

s. xiii¹

Arendal, AA 4981-2 Oslo, MCH, C 34738/52

Seqv?

2

3

Unkn.

s. xiii²

roll

-

1

Norway

s. xiii m.

Seqv 60 add Seqv 64 add

Cph. KB NKS 133 f. 4 NRA, lat. fr, 1030

Man

-

3

Norway

Ant

2

1 (inc)

Norway

s. xiii (and xiv) s. xiii²

Seqv 39b add, Mi 38 Seqv 42 add

Found in Lom stave church

Nordlands len 1610

Purely by chance three out of the four liturgical once-complete manuscripts with Norwegian scribes listed by Lilli Gjerløw in her edition of the Nidaros ordinal (Gjerløw 1968, 34-8) are represented among the manuscripts containing sequences. The first is a missal written in correspondence with the Nidaros ordinal (Seqv 39b add/Mi 38), by a scribe who also wrote an antiphoner and who has some connection to the Old Norwegian Homily Book, which will be discussed in the first chapter. The second scribe wrote a breviary-missal, also in correspondence with the Nidaros ordinal (seqv 38 add/Br-Mi 3), with a secondary provenance in Stavanger. I would like to suggest that Seqv 38 add is a manuscript from western Norway, and I compare it with a missal of the same size and general lay-out, presumably from Nidaros. The third manuscript from Gjerløw’s list is written by a scribe known as the St. Olav scribe. In the following chapters the manuscripts will be treated chronologically, beginning with the first manuscript written secundum ordinem (Seqv 39b add/Mi 38).

156

5.1. The missal and the Old Norwegian Homily Book: Seqv 39b add (Mi 38), s. xiii¹

5.1. The missal and the Old Norwegian Homily Book: Seqv 39b add (Mi 38), s. xiii¹ This first section about the thirteenth century manuscripts is devoted to a missal primarily labelled by Lilli Gjerløw as Mi 38 (Seqv 39b add), of which only one leaf is left (Oslo, NRA Lat fragm. 764). The leaf, which measures c. 23 x 14,5 cm, is cut in half and slightly cropped on both sides. Mi 38 has a significant sequence incipit, making it the first missal known to be written in accordance with the Nidaros ordinal (cf. Gjerløw 1968, 34-5). Because of its scribe the missal is connected to two other manuscripts, a fragmentary antiphoner (labelled Ant 7) and the Old Norwegian Homily Book (Copenhagen, AM 619 4°), the oldest remaining Norwegian codex in the vernacular. In the following chapter the date of the missal will be discussed, its scribe and the connection to the other manuscripts, and most importantly, if the three manuscripts seen together can bring us close to a common place of origin.

5.1.1. An unusual sequence Mi 38 was presented by Lilli Gjerløw in an article on missals used in the bishopric of Bergen (Gjerløw 1970, 109-11). The rest of the leaf contains chants without notation, texts and prayers for the third Sunday of Advent to the feria sexta, Friday, of the third week of Advent (for a detailed description of the contents, see Gjerløw 1970, 110-11). The sequence, indicated with incipit only, for the third Sunday of Advent is the unusual Gaudia mundo ventura (AH 37, no. 1) rather than the commonly used Qui regis sceptra (AH 7, no. 9 and AH 54, no. 3) for this Sunday. The sequence Gaudia mundo is not otherwise known to be used for the third Sunday of Advent (cf. table in Ommundsen 2006b, 146). Three French sources from the fourteenth century, among them a St. Denis manuscript (Paris, Bibl. Nat. 1107), have Gaudia mundo for the first Sunday of Advent, and the early fifteenth century Sherbourne Missal (London, British Library Add. 74236) uses it for the fourth Sunday of Advent (Gjerløw 1970, 110). The presence of the Gaudia mundo, a distinctive and rare feature, is a strong indication that our missal was following the Nidaros ordinal, making Mi 38 the oldest missal fragment secundum ordinem (Gjerløw 1968, 35; 1970, 110; 1979, 242).

157

Part II: Analysis Fig. 14: Oslo, NRA, Lat. fragm. 764 (Mi 38/Seqv 39b add). The sequence Gaudia mundo ventura is indicated with incipit only.

The two halves of the missal leaf were used to bind an account from Bergenhus len in 1612. Although it is problematic to determine provenance with fragments from only one account, Bergen in this case is likely as a secondary provenance, as was also suggested by Gjerløw (1970). Lilli Gjerløw gave a terminus ante quem non at c. 1225 for Mi 38 (Gjerløw 1970, 109), presumably guided by the strictest terminus ante quem for the Nidaros ordinal, which she originally set to 1224 (Gjerløw 1968, 31). In chapter 1.4.3. above the discussion of the date of the ordinal concluded that the period 1202-13 was the most likely timeframe for the completion of the ordinal. It is therefore likely that the missal was copied after 1200, but not necessarily as late as after 1225. In fact, a date in the first quarter of the thirteenth century seems likely, due to its palaeographical features and the circumstantial evidence of the other manuscripts.

5.1.2. Two liturgical manuscripts – one scribe Lilli Gjerløw identified the scribe of Mi 38 as the same scribe who penned the antiphoner Ant 7 (Oslo NRA, Lat fragm. 1018, 1027, 1034 and an unnumbered fragment) (Gjerløw 1968, 35). The antiphoner labelled Ant 7 was edited and described by Gjerløw in the Antiphonarium Nidrosiensis Ecclesiae (1979, 242-50). It is preserved in twenty loose leaves of which six are whole with a size of 30 x 20 cm. The writing space is c. 23 x 15 cm in one column, with 12 lines to the page (with staves). The initials are in red and green (with penned decoration, green on red and vice versa), 2 lines high. The script-samples of the missal and antiphoner provided below show some of the significant features of the scribe, and a comparison makes it clear that Gjerløw was right in assigning the two manuscripts to one scribe. Notice the eight-shaped g in

158

5.1. The missal and the Old Norwegian Homily Book: Seqv 39b add (Mi 38), s. xiii¹ both antiphoner and missal, the tall and sometimes trailing-headed a, and the characteristic shape of the uncrossed tironian note for “et”. Other features typical for this scribe are the t beginning with a hairline or approach-stroke and the hyphens beginning at the baseline and holding an angle of approximately 45°. The shape of the g and a, and the dotted y’s, attest to English influence on the script, which also corresponds well with the shape of the punctus elevatus. The script may be characterized as Pre- or Protogothic, since biting does not extend beyond two consecutive p’s.

Fig. 15: Above: Oslo, NRA, unnumbered leaf, Salten 1635 (Ant 7), two lines from St. Olav’s office In regali fastigio. Below: Oslo, NRA, Lat. fragm. 764 (Mi 38). On comparison is seems clear that the antiphoner and missal are written by the same scribe, as pointed out by Lilli Gjerløw.

Gjerløw refers to the Ant 7 as “the St. Edmund Antiphoner” on account of the presence of the proper Office of St. Edmund. Apart from the St. Edmund Office, which is not laid out in detail in the Nidaros ordinal but prescribed merely as the “propria hystoria” (proper Office) of St. Edmund (Gjerløw 1968, 410), only two of the nineteen Offices in Ant 7 conform wholly to the Nidaros Ordinal, namely those of St. Olav and SS. Peter and Paul. The others have variations in the contents and in the order of different antiphons or responsories compared to the ordinal, leading Gjerløw to ask if Ant 7 represents an early version of the antiphoner of the ordinal or a stage in the editorial process (Gjerløw 1979, 250). Gisela Attinger, in her thesis on the chant melodies of Nidaros antiphoners, concluded that although Ant 7 does not conform wholly to the Nidaros ordinal, it is “clearly related to Nidaros not only with respect to the selection and order of text, but also melodically” (Attinger 1998, 282).

159

Part II: Analysis The leaves from the antiphoner were used in bindings from northern Norway (Nordlands len, Tromsø and Salten) in the period 1625-1648. The fact that one manuscript was used for bindings from the same fief for such a long period of time makes it likely that the secondary provenance of Ant 7 was northern Norway. Although Gjerløw suggested that the antiphoner represents an early stage of the ordinal, she proposed a date within the second third of the thirteenth century (Gjerløw 1979, 242), i.e. after 1230. Again, there do not seem to be anything in either script or contents for such a late date. Quite the contrary, if the contents suggest that the antiphoner was written around the introduction of the Nidaros ordinal, and perhaps even before it had its final form, a date in the first quarter of the thirteenth century is more likely.

5.1.3. One scribe or many? The problem of the Old Norwegian Homily book The third manuscript connected to the missal and antiphoner is the Old Norwegian Homily Book (AM 619 4°). It has been dated c. 1200 or 1200-1225,154 and is thereby the oldest Norwegian book in the Old Norse language. The homily book is a codex of 80 leaves, 23 x 15,5 cm. The text is written in one column with 30 lines to the page. The initials are multicoloured and quite elaborate in red, green and blue. The contents are c. 30 sermons in Old Norse for the course of the church year. Four leaves are the remains of another book of similar type, inserted into the manuscript at an unknown date. The initials on these leaves are in red, green and brown. The place of origin was early identified as the northern parts of western Norway (i.e. Bergen or the monastery Selja at Stad), based on linguistic indicators (Hægstad 1906, and by Knudsen in Seip 1952). The sermons in the current homily book seem to be copied from one or more older books, as indicated by the number of archaisms in the language. The identification of western Norway as a place of origin for the book is corroborated by the mention of “í Þrondhæimi norðr,” i.e. north in Trondheim, and “i Vic austr,” i.e. east in Vik (present day Oslo) (Hægstad 1906, 75). 154

For a survey of the discussion regarding dating, see Indrebø 1966 (1931), 38.

160

5.1. The missal and the Old Norwegian Homily Book: Seqv 39b add (Mi 38), s. xiii¹

While Unger, its first editor, suggested that the book, including the four inserted leaves, was written by one scribe, later scholars have identified two scribes for the book and one for the inserted leaves (Hægstad 1906, 41) or two hands in the book itself and two on the inserted leaves (Indrebø 1966 (1931), 69-72). This opinion has been generally accepted since then. The distribution of the hands according to Indrebø is: Hand 1: ff. 1 - 16, and f. 78 line 26 - f. 80 Hand 2: ff. 17 - 68 and f. 73 - f. 78 line 25 Hand 3: f. 69 – f. 71 line 27 (inserted leaves) Hand 4: f. 71 line 28 - f. 72 (inserted leaves) This view is now being contested. Michael Gullick has studied the script of the Old Norwegian Homily Book, and states that instead of four, the manuscript, including the inserted leaves, is the work of one single scribe (Gullick forthcoming). He also states that the same scribe wrote the antiphoner and missal (ibid.). While Gjerløw pointed out that the scribe of Mi 38 seemed to have been formed in the same school of script as the scribes in the Old Norse homily book, she did not venture further than to suggest familiarity, not identity (Gjerløw 1968, 35). This development is an interesting parallell to the research on the scribes in the Icelandic Homily Book (Stockholm, Royal Library, Perg. 15 4°). From going from one scribe in the first edition (Wisén 1888), a varying number of scribes have been identified culminating in fourteen and twelve scribes in two thesises from the mid seventies (Rode 1974; Westlund 1974). In the latest edition the suggested number of scribes is again reduced to one (Leeuw van Weenen 1993). 155 The tendency in the evaluation of scribes in the later years seems to be a better understanding of the different variables which may influence a scribe’s handwriting: different book genres may require different levels of formality, a scribe could change his or her style depending on new scribal impulses from abroad, the influence from his or her 155

I am grateful to Jonas Wellendorf for making me aware of the parallel between the latest development in the Old Norwegian Homily book and that of the Icelandic Homily Book.

161

Part II: Analysis exemplar, the requirements of a given employer, the equipment or even the dayform. Before moving on the question of origin, it is necessary to investigate, first, if the Homily book really is written by one person, and second, if this person is the same who wrote the antiphoner and missal. The following script-samples are from the facsimile-edition of the homily-book:

Fig. 16: “Hand 1” of the Homily book. Above f. 2r, below f. 16r

Fig. 17: “Hand 2” of the Homily book, f. 36r.

“Hand 1” is illustrated with two samples to demonstrate the growing informality of the scribe’s hand between the leaves 2 and 16, showing itself in the tall a’s and the documentary-style g. When comparing the “hand 1”of the first leaves with “hand 2” the hands look remarkably similar, particularly the shape of the uncrossed tironian note for “et” and the shape of the g. Both also have hyphens starting on the baseline, rising in a similar angle. In my opinion “hand 1 and 2” are in fact one scribe, and it seems clear that the original Homily book (without the insertions) was written by

162

5.1. The missal and the Old Norwegian Homily Book: Seqv 39b add (Mi 38), s. xiii¹ one scribe. Although the scribe of the Homily book is a fairly good scribe, he is presumably writing fairly quickly, and there is a variation in the shapes of the letters. The former palaeographical descriptions are therefore about tendencies – whether the a tall or the g are written rather “formally” or with a “documentary-loop”. Differences or “tendencies” regarding formality do not seem to be significant when separating one scribe from another, especially not when the general appearance is as similar as in this case. The samples from the inserted leaves do not display the same general steady quality as the ones above:

Fig. 18: “Hand 3 and 4” of the Homily book, on f. 71v. The break between the hands is set after the rubric, at the beginning of the new sermon.

The quality of the script on the inserted leaves immediately looks less assured, although it is basically the same type of script. On the first leaf “hand 3” looks more like the scribe responsible for “hand 1 and 2” (cf. illustration below). The general appearance of “hand 3” and in “hand 4” is much the same, and so are also the loops of the g’s. The “break” identified on f. 71v seems to have more to do with the quality of ink and pen than the scribe. To me it is also seems clear that “hand 3 and 4” are the same scribe. Now, as to the question whether the scribe of the original Homily book and the scribe of the inserted leaves are the same person. It seems peculiar to me that one scribe should be able to keep up such a good general quality in the course of a whole book, and then show such variable style in leaves from another book, even though it seems clear that in these leaves he had some problems with ink, pen, parchment or all three. One could argue that the inserted leaves were in fact from a preliminary copy or version of a homily book, but then it would seem strange that it contained decorated initials. I am reluctant to identify the scribe of the Homily book 163

Part II: Analysis as the scribe who wrote the inserted leaves, although I would also be reluctant to exclude it as a possibility. How do the scribe or scribes of the Homily book relate to the liturgical fragments? The script in the liturgical books is basically the same, but perhaps slightly more constrained (to avoid using the word “formal”). The two liturgical manuscripts do not display the “loose” quality that the scribe of the Homily book lapses into at times, and the g has a “tighter” appearance, especially in the missal, making it to a larger degree resemble the number 8. The majority of the a’s in Mi 38 and Ant 7 seems to be trailing-headed and with a “squarer” body than the ones in the Homily book. This could possibly be a result of the slightly more formal genre the liturgical books represent. It could also be the case of a scribe changing his hand a little over time. In conclusion, I agree with Michael Gullick that the scribe of the Homily book also wrote the liturgical manuscripts Mi 38 and Ant 7. The scribe of the inserted leaves may very well be the same, but here a firm conclusion is more difficult to make.

5.1.4. A single scribe or a scriptorium? Is it possible to show if these three books (or four, counting the inserted leaves of the Old Norwegian Homily Book) were just the products of one scribe working alone, or is there evidence of a scriptorium in these works? For this it is possible to turn to the codicological evidence, especially the initials. The size and proportion of Mi 38 is remarkably much like the Homily book. They both measure c. 23 x 15 cm. The antiphoner is somewhat larger, c. 30 x 20 cm. The pigments look fairly similar. The red in Mi 38 is of good quality, but the green has a tendency to “bleed” into the parchment. In the Homily book the green has sometimes even eaten through the parchment, and would appear to be of the same type (although a comparison of hue has not been perfomed). The scribe wrote around pre-existing holes in the parchment both in the missal and in the Homily

164

5.1. The missal and the Old Norwegian Homily Book: Seqv 39b add (Mi 38), s. xiii¹ book. The nature of the parchment would probably say the most about the place of production, but this feature remains to be investigated. The red and green initials in Mi 38 and Ant 7 with their simple penned decorations form a contrast to the elaborate initials in the Homily book. However, the plain initials of the liturgical books do not have the “homemade” quality of many other initials among the Norwegian sources. Looking at them more closely, they all seem to display similar features: The coloured initials are from the missal, the ones in black and white are from the antiphoner. The two h’s, apart from being a similar shape, both have a verticle line dividing the counter space of the letter. Two other initials, the I from the missal and the M from the antiphoner, show an additional feature, namely a particular “curl” at the end of a line resembling a scorpion’s tail. Instead of simply ending in a “dot”, the line ends in a dot with a hairline resembling a stinger. Fig. 19: Left (colour): Oslo, NRA, Lat. fragm. 764 (Mi 38/Seqv 39 add). Right (bl&wh): Selected initials from the antiphoner Ant 7 (illustrations taken from Gjerløw 1979, Pl. 63-64)

Based on a simple comparison of shapes and features, it is possible as a first step to draw the conclusion that not only the text, but also the initials of the missal and the antiphoner most likely were done by the same person. The next question is if this is the same person who made the initials in the Old Norwegian Homily Book. In spite of the letters in the liturgical books being simpler (the remaining ones, at least), the presence of the same features, the line dividing the rooms of letter, the scorpion’s tale and the frill, could indicate that the artist was one and the same for all four books: the missal, the antiphoner, the Homily book, and the inserted leaves.

165

Part II: Analysis Whether the artist doing the initials was the main scribe, is not easy to say. The distinct features occurring in all the manuscripts, like the dividing line and the scorpion’s tale, could be a typical for the scriptorium and not just one scribe. Michael Gullick has identified the scorpion’s tale in other fragments, relating them to this group (Gullick, forthcoming). He was also the first to identify another significant feature of the initials of the Homily book, namely the “split petal” (resembling a coffee bean), visible in the round M above. This particular feature is generally connected with books produced in northern England (Gullick 1998, 259).

Fig. 20: The M to the right is from the main part of the Homily book. The presence of the “split petal” (resembling coffee beans), first identified in the Homily book by Michael Gullick, indicates a northern English influence. Fig. 21: The M to the left is from the inserted leaves in the Homily book. It contains some of the features from the M in the antiphoner, such as the line dividing the upper mid space, the “frills” on the left stem, and the scorpion’s tails (one in each space below the arches of the letter).

That the initials are in the same style throughout the Homily book, although the sermons were most likely collected from different sources, is in favour of the initials representing the style of the artist or the scriptorium where they were made, and that they have not been copied from the exemplar. In the centuries preceding the Homily book, at least, initials could be regarded as the “signatures” of a scriptorium: Parkes points out that although a manuscript could be copied by hands exhibiting wide

166

5.1. The missal and the Old Norwegian Homily Book: Seqv 39b add (Mi 38), s. xiii¹ regional differences “the scriptorium tradition seems to survive in the initials” (Parkes 1982, 23). The fact that we have remaining evidence of four books with very similar script, possibly all by the same scribe, and initials with some of the same features, (although some simple, some with elaborate decoration) indicates that the books were produced in one place with a good supply of parchment and pigments, not on several different locations (although careful codicological examination and comparisons, for instance of the parchment, has not been done). So when Ant 7 ended up in northern Norway, it was probably brought there in the form of a book – not as a result of the scribe himself travelling and producing books in Nidaros or northern Norway as well as in Bergen (Gjerløw 1979, 242). It could either have been brought via Nidaros by a priest, or it could have come directly from Bergen in connection with the frequent traffic northward from Bergen at some point before the Reformation. The scriptorium was probably in one of the religious institutions in Bergen, rather than being a private workshop, as such an establishment seems unlikely in Norway as early as the first quarter of the thirteenth century. The scribe of the missal, antiphoner and Homily book must have had a steady production of manuscripts, most of which are probably lost, taken into account the generally poor transmission of medieval manuscripts in Norway. Both the features of his script and the split petal motif in the initials suggest that he worked in a place under English influence. The three manuscripts (or four, counting the inserted leaves) can be seen as part of the same historical context. Together they form the impression of an environment where books were produced both in Latin and Old Norse, at a time when a scribe producing liturgical manuscripts to an increasing degree had to relate to an issued ordinal for the archbishopric. This group of manuscript seen together can possibly bring us closer to the place of origin.

167

Part II: Analysis 5.1.5. Suggesting an origin As mentioned above the linguistic features of the Old Norwegian Homily Book places it in western Norway or Bergen, which is supported by Mi 38 having a likely secondary provenance in Bergen. The scholars studying the Homily book have made several suggestions for a more specific place of origin, like Munkeliv abbey in Bergen or the Benedictine monastery at Selja (Hægstad 1906, 75). Regarding Selja, the latest notion has been that such a fine book with multicoloured decorations was most likely not produced there, since Selja by 1200 had lost its position as formal bishop’s see to Bergen. In the preface to the facsimile edition of the Homily book (Seip 1952) Trygve Knudsen refers to Mikjel Sørlie (1950), who sets two possible alternatives for the scriptorium in question, both in the centre of Bergen: 1) The Cathedral Chapter (northern side of the town bay) 2) The Benedictine monastery of Munkeliv (southern side of the town bay) Of these two suggested institutions one is secular and one is monastic. Elements in the contents of the Homily book have been taken to point in the direction of a monastic setting. The sermon “In die pentecosten” contains references to a monastic way of life: “Vér hofum allt fyrir-látet fyrir guðs sakar, bæðe fe ok frændr, ok omnem mundi pompam. Vér haofum ok omnia communia sem umm þa er leset” (We have forsaken everything for God’s sake, both property and family, and all the splendour of the world. And we also hold all things in common, as is read about that). The sermon does not refer to “monks” directly, but mentions life in the cloister (í claustra) (cf. Indrebø 1966 (1931), 59, 92, 94). Whether liturgical books are secular or monastic is in some cases possible to find out. While Mi 38 could be either secular or monastic, the antiphoner Ant 7, on the other hand, show clear signs of being produced in and for a secular, or at least nonBenedictine, setting: the Offices of Ant 7 are presented with three antiphons of the nocturns at matins, which reveal that the antiphoner was not made for Benedictine use. A Benedictine community would recite twelve psalms at matins, i.e four psalms in each nocturn, requiring four antiphons and responsories. One may of course imagine that Munkeliv had been assigned the task of writing the antiphoner with specific requirements from a secular employer, and that it was never intended to be 168

5.1. The missal and the Old Norwegian Homily Book: Seqv 39b add (Mi 38), s. xiii¹ used in the monastery. There is one feature making that scenario quite unlikely, namely that the feast of St. Benedict’s translation (11 July), generally celebrated also in the secular liturgy, is excluded (Gjerløw 1979, 243). The period between the Commemoration of St. Paul (30 June) and St. Peter’s chains (1 August) remains more or less in full, and it is therefore not a question of Benedict’s Office being on a lost piece of the book. As the Nidaros ordinal, which was just around the corner at this time, prescribes the translation of St. Benedict, it would be odd if a Benedictine scriptorium would not enter this feast in the antiphoner. The liturgy for the translation is not the same as that for the natale (21 March), and thus cannot be explained by the availability of the Office elsewhere in the antiphoner. The conclusion regarding Ant 7 is that it was written for use in Norway (with the St. Olav Office), but not in or for a Benedictine monastery. To sum up the evidence so far, we have: 1) A missal secundum ordinem (could be monastic or secular). 2) An antiphoner, close to, but not in strict accordance with the Nidaros ordinal – probably written in a non-Benedictine scriptorium for use in a non-Benedictine church. 3) A Homily book in Old Norse – presumably written in and for a monastic context. The situation so far seems to be in a deadlock: While the Homily book points towards a monastery, Ant 7 points away from it. There is, however, one other alternative and possible solution, namely that the scribe worked in a monastic community which did not follow the rule of St. Benedict. In the preface to the translation of the Old Norwegian Honily Book Erik Gunnes suggested that the transmitted copy or redaction of the homily book could have been made either in St. John’s house of regular canons or at the bishop’s see in Bergen (Gunnes and Salvesen 1971, 16-17), a notion supported by Knut Helle (1993, 118). In his commentary Gunnes discusses the origin of the sermon “In die pentecosten”, and points to possible allusions to the rule of St. Augustine. The sermon (cf. quotation above) resembles the rule of Augustine (cap. 1) on property: “Sic enim legitis in Actibus Apostolorum, quia erant illis omnia communia...” (For this you read

169

Part II: Analysis in the Acts of the Apostles, that they held all things in common) (PL 32, 1379). Gunnes points out that this particular sermon is not in the Icelandic Homily Book, and does not have linguistic features placing it among the older layer of sermons in the Homily book. As the language in this sermon is influenced by south-western Norwegian, Gunnes suggests that either St. Olav’s in Stavanger or Halsnøy south of Bergen could be the sermon’s primary place of origin (1971, 174-75).156 Oddmund Hjelde, who has studied the literary sources of the sermons in the Homily book and their intertextual relationships, supports Gunnes’ conlusion that the pentecost sermon has its origin in one of the Augustinian houses in western Norway (Hjelde 1990, 288). The question is if not only the pentecost sermon, but also the Homily Book itself can be linked to an Augustinan house. Although the houses further south cannot be completely disregarded, the house that seems the most likely is St. John’s in the heart of Bergen. St. John’s was certainly founded by 1208, but probably before 1180. Olav Nenseter, who wrote his thesis on the Augustinians in Norway, suggests that St. John’s could have been founded as a collaboration between archbishop Eystein and Erling Skakke and/or his son Magnus, who was crowned in Bergen in 1163/64, some time before archbishop Eystein’s exile in England (1180-83) (Nenseter 2003, 52-53). While the Augstinian houses in the south-east of Norway are assumed to have been under Victorine influence, the Augustinian houses on the west coast seem to have been influenced from England (Nenseter 2003, 32-33). St. John’s was small and not wealthy, and probably had less than twelve residing canons. Still, it was centrally placed in the town of Bergen with is own pier in the town harbour, close to Munkeliv on one side and with the bishop and the secular canons just across the bay. Sverre’s saga accounts that when archbishop Eystein came back from England in spring 1183 his ship laid at St. John’s pier (Unger and Vigfússon 1862, 603-604). Later that same summer he went north to Trondheim. Although the house itself is not mentioned, it would seem that if a house of regular canons really was established at that time, this house was the archbishop’s chosen place of residence (Nenseter 2003, 51-52). So, 156

For later discussions regarding the complexity of regional linguistic features in Old Norse texts, and to what degree they can be ascribed to the exemplar or the scribe, see for instance Hødnebø (1984), particularly pp. 170-71.

170

5.1. The missal and the Old Norwegian Homily Book: Seqv 39b add (Mi 38), s. xiii¹ although St. John’s was not wealthy, it had powerful protectors and was centrally placed in an important town, and it is likely that some of the canons would produce books of a quality like the Old Norwegian Homily Book. When the Homily Book is not regarded separately, but in connection with the liturgical books, this strengthens the argument for St. John as a possible place of origin. An Augustinian house like St. John’s could account for the “monastic, but not Benedictine” context indicated by the homily-book and antiphoner. The remaining evidence of the four books points in the direction of an institution where a scribe could work under relatively stable conditions, with a steady supply of parchment (although some holes had to be accepted), a place under English influence as shown through both script and initials, a place with a close connection to the archbishop (if the Ant 7 is in fact, as suggested, an “early redaction” of the Nidaros ordinal), a monastery, but not a Benedictine monastery. It seems worth putting forth as a hypothesis to be investigated further in future researh that the Old Norwegian Homily Book (AM 619 4°), the missal Mi 38/Seqv 39b add (Oslo, NRA Lat. fragm. 764) and the antiphoner Ant 7 (Oslo, NRA Lat. fragm. 1018, 1027, 1034 a.o.) could be the products of a regular canon working in the Augustinian house of St. John in Bergen.

171

Part II: Analysis

5.2. On the track of an unknown local scribe: Seqv 51 add, s. xiii¹ A manuscript of unknown genre labelled Seqv 51 add (Oslo, NRA, Lat. fragm. Sandaakers katalog LR pk. 157, Bratsberg) is represented by one leaf, which is transmitted more or less undamaged. It contains the middle part of the sequence Iocundare plebs fidelis (AH 55, no. 7) for the evangelists, which is a sequence not prescribed in the Nidaros ordinal. Two more leaves have been identified with the same scribal characteristics, but the leaves may come from different books. If the leaves are from more than one book, we may be on the track of a hitherto unknown Norwegian scriptorium.

5.2.1. Scribal characteristics The scribe who wrote this manuscript in the first decades of the thirteenth century is somewhat of a mystery, because it is easier to say where he or she was not from, than to link him or her to one particular European region; Some features, like a tall a, look English, while other features, like the shape of the ampersand, points away from England and possibly towards a French influence. The “Anglo-French” shape of the g, with a ductus influenced by documentary script, indicated that he or she was not German. After looking through a large number of reproductions of Norwegian fragments, kindly provided by Michael Gullick, it soon became clear that the single fragment labelled Seqv 51 add was not the only place these scribal features appeared. A fragment from a lectionary-missal labelled Lec-Mi 4 in Lilli Gjerløw’s catalogue was written by a scribe with the same characteristics. In this chapter I will ask the following questions: First, is the scribe of Seqv 51 add the same scribe as that of LecMi 4? Secondly: are Seqv 51 add and Lec-Mi 4 actually from the same book, and how was this book organized? A third fragment has also been identified, this time from a gradual, labelled Gr 27, with some of the same characteristics as the other two. How is Gr 27 related to the other two? And if the three mentioned leaves are in fact evidence of different books, where could they have their origin?

172

5.2. On the track of an unknown local scribe: Seqv 51 add, s. xiii¹ Fragments from the lectionarium missae Lec-Mi 4 (Oslo, NRA Lat. fragm. 452) were discussed at a workshop in Oslo in August 2006. The features of the script and the quality of the parchment made the workshop participants believe that Lec-Mi 4 had a Norwegian origin, and it was datable to c. 1200.

157

Lec-Mi 4 consists of three cropped

leaves. The contents are lessons from liber sapientiae, for the epistle-reading at mass. What is most interesting is the scribe. Notable are the shape of the ampersand and the tall a, and the ductus of the Anglo-French g influenced by documentary script, all features reminiscent of Seqv 51 add. Furthermore there are some other significant features indicating a very close relationship of the hands of Seqv 51 add and Lec-Mi 4: -

the sharp angle of the e and c

-

the “dip” on top of the letter q, and the sharp upward-going serifs added to the descenders of q and p.

-

the use and shapes of the letter d, which is both round without top-stroke, round with top-stroke and straight (also with a “dip”).

-

the ductus of round r after o, with the final line in the r being almost vertical, with a starting point far to the left. See Seqv 52 add, the left column on the verso side of the leaf (Part II, Catalogue). See Lec-Mi 4 below, second line from below, left column.

-

The shape of the suspensions sign for omitted m and n: A horizontal line with vertical stroks on the sides.

The nib of the pen used to write the leaf of Seqv 51 add is smaller than the one used for the text of Lec-Mi 4, which accounts for some of the differences in the general impression of the two pages. The “neater” look of the leaf of Seqv 51 add may be due to this thinner pen and the “organizing” effect of the staves on the page. The rubrics in Lec-Mi 4, however, were written with a pen of the same thin type used for the lyrics in Seqv 51 add. The ink in Seqv 51 add is a quite pale light brown, and there is a “see-through” effect in the parchment which makes it more difficult to read. The 157

I am grateful to Michael Gullick for showing me the material from this workshop, and for discussing with me the possibility of a common scribe in Seqv 51 add and Lec-Mi 4.

173

Part II: Analysis parchment of Lec-Mi 4, on the other hand, is quite thick and opaque.158 In spite of the differences in pen-nibs and parchment it seems clear from the particularities of the script-samples above that the scribe in Seqv 52 add is in fact the also the scribe of Lec-Mi 4. Fig. 22: The small segments to the left: Seqv 51 add (colour) and Lec-Mi 4 (bl-wh). Notice the similar shape of the d’s and the e’s

Fig. 23: To the right: Segment of Seqv 51 add (Oslo, NRA, Lat. fragm. Sandaakers kat. LR, pk. 157). Notice the ampersand, the tall a' s, the shape of the g and the q.

Although we can say that the scribe of Seqv 51 add and the scribe of Lec-Mi 4 is one and the same, it cannot be automatically assumed that the leaves come from the same book. The reproductions of the fragments show clear indications that they could come from the same book, primarily on grounds of the format and lay-out of the pages. Seqv 51 add measures c. 31 x 21 cm, while the leaf of Lec-Mi 4 measures 29 x 16 cm, cropped in the lower margin and on both sides, although its original size must have been very close to that of Seqv 51 add. When comparing the writing space, Seqv 51 add measures c. 22 x 16 cm, while Lec-Mi 4 has a writing space of 21,5 x 16,5 cm, which are fairly similar.

158

I thank Michael Gullick for this observation.

174

5.2. On the track of an unknown local scribe: Seqv 51 add, s. xiii¹

Fig. 24: A copy of Lec-Mi 4 (Oslo, NRA, Lat. fragm. 452): Notice the peculiarities of the scribe, particularly the special ampersand, with the thick upper lobe, and the tall a' s. Notice also the difference in the width of the left and right columns. The reproduction is not in 1:1.

175

Part II: Analysis

Two more points can be remarked on the lay-out: The pricking of Seqv 51 add shows that even though the page has 22 lines with staves, the leaf was initially prepared for 26 textlines. Although the pricking is lost in Lec-Mi 4, the text is written on 26 lines, which correspond with the prickings in Seqv 51 add. A final formal feature is the difference in width of the two columns in both leaves: the inner column is 1-2 cm wider than the outer column. The format and lay-out of the two leaves appear to coincide, and certainly does not ecxclude that the two leaves came from the same book. The available circumstantial evidence may provide additional information: when and where were the fragments in Seqv 51 add and Lec-Mi 4 used to bind accounts? Seqv 51 add was found on taxaccounts from Bamble (Bratsberg) 1611-13. The fragments labelled Lec-Mi 4 were used on bindings from Bratsberg 1651. The accounts come from the same fief, although almost 40 years apart. Is it reasonable to assume that a dismantled book was laying around a bailiff’s office for that long? Another question worth asking is: would a lectionary for the Mass even include sequences? To celebrate mass the books are needed to provide 1) the prayers 2) the lessons 3) the liturgical chants and 4) the sequences. Initially a sacramentary would provide the prayers, a lectionary would contain the lessons for the epistle and gospel-readings, a gradual the chants, and a sequentiary would provide the sequences. A missale plenum would contain all these elements in one book. Still – a missal could be organized in different ways: it is not unusual for sequences to be found assembled in the back of a missal instead of being spread out according to the church year. At first glance it does look like the lectionary here has the lessons in continuous order – not mingled in with the chants. That opens up two possibilities: the original book contained the lessons for the mass only, or, what one could also imagine, it was a book containing all the elements of the mass separately in different sections.

176

5.2. On the track of an unknown local scribe: Seqv 51 add, s. xiii¹ 5.2.2. A third manuscript This brings us to the third manuscript in question, namely a gradual which Lilli Gjerløw labelled Gr 27 (Oslo, NRA Lat. fragm. 433), also used to bind an account in Bratsberg, this time for the year 1652. The feature that drew attention to it in connection with Seqv 51 add was not initially the script, but the notation. Gr 27 contains the same kind of “flat” notation as Seqv 51 add. To find notes as “horizontally prolonged” as this is not very common (and when they do appear it is generally in fragments believed to be local, see for instance Seqv 22b and Seqv 60 add). And although the scribal features isolated above seem a little less pronounced here, some of them are in fact present: -

the “dip” on top of the q, and the serifs added to the descenders of q and p

-

the sharp angle of the c and e

-

the ductus of the g influenced by documentary script

There are no good examples of tall a, and no ampersand or straight d on the available reproduction. What does appear in Gr 27 is an æ-ligature, in the name Mælchisedech, which seems odd for a gradual at this time. An æ-ligature after the year 1200 may be seen as an influence from the vernacular, and consequently another indication that the scribe is Scandinavian or Norwegian. Let us first briefly ask: is the scribe in Gr 27 the same as the one we have just been acquainted with? The script on the gradual leaf certainly resembles that on the others to such a degree that the most likely conclusion is that one scribe wrote the leaves of all three “manuscripts”, the sequentiary, the lectionary and the gradual. While the other fragments displayed prose text with a thick pen, and rubrics and chant text with a thin pen, Gr 27 is a third setting, namely chant text written with a thick pen. The thicker pen could have the effect of the script looking more “compressed”. What seems certain in any case is that if Gr 27 was not written by the same scribe, it was written by someone trained in the same place, working alongside a music scribe trained in the same place as the one in Seqv 51 add.

177

Part II: Analysis

Fig. 25: Gr 27 (NRA, Lat. fragm. 433). The flat notation resembles that of Seqv 51 add, and so does the hand. Notice the p' s and q' s and the shape of the e. Notice also the æ-ligature in Mælchisedech, four lines up in the right column.

178

5.2. On the track of an unknown local scribe: Seqv 51 add, s. xiii¹ Since there are features in the script and notation linking Gr 27 to Seqv 51 add and Lec-Mi 4, we should not move further without asking if Gr 27 was once a part of the same book as Lec-Mi 4 and Seqv 51 add. To look at the format first, the cropped page measures c. 28,5 x 17,5 cm, i.e. it was originally a bit taller and a bit wider, which corresponds well with the format of the other two. The writing space is c. 23 x 16*, which is c. 1 cm taller than the other two, but still relatively close. The number of lines with staves are 22 – the same number of lines as in Seqv 51 add. The two columns, however, seem to be of equal width, and not of different widths like in Seqv 51 add and Lec-Mi 4. Furthermore the general aspect of the small initials on the page do not seem to resemble either Seqv 51 add or Lec-Mi 4. The leaf from the manuscripts labelled Gr 27 was used on a tax-account dated 1652, which is close to the date on the account of Lec-Mi 4, 1651. The years on the accounts could indicate that these two leaves were bound in the same codex, at least at the time of their reuse. Going by the dates, Seqv 51 add, used as early as 1611-13, was probably not bound together with the other two leaves. The consequences are striking and significant, almost regardless of which conclusion is drawn: either we have the remains of one “composite” book with a lectionary, a gradual and a sequentiary, or we have the remains of two or three books with the same scribe/scribal features, which would point us in the direction of a local scriptorium. The alternative conclusions are listed in fuller detail below: Alternative 1: We could be dealing with a missal with separate parts: a lectionary, a gradual and a sequentiary (and possibly a sacramentary), in other words all the texts and chants necessary to celebrate mass gathered together in one book, but in the “old-fashioned” way – not the “modern” missale plenum. If I am right in assuming that the scribe is the same in all three fragments, we would be dealing with one interesting book and one interesting scribe – but sadly the contours of our “scriptorium” would become less clear.

179

Part II: Analysis Alternative 2: It is possible that Seqv 51 add, Lec-Mi 4 and Gr 27 all come from separate books. That would be a truly interesting scenario, because then we have three books written by the same scribe re-used for binding accounts in one fief, presumably collected in the same region. In that case we would be closing in on a local scriptorium. What would first come to mind is the female Benedictine house of Gimsø, which was situated in what later became Bratsberg len. At this point this can be nothing more than a speculation, but a closer investigation into this matter could start by examining more bindings from Bratsberg in search of more fragments written in the style of Seqv 51 add, Lec-Mi 4 and Gr 27. Alternative 3: Lec-Mi 4 and Gr 27, both used in bindings c. 1650, could be from the same book, and Seqv 51 add from another. That would leave us with a situation not very different from the one above, as there would be one scribe present in more than one book. As previously mentioned a possible future project could involve a search among the other fragments used on bindings in Bratsberg for the same scribe or sribes writing with similar features. This might bring us closer to a local scriptorium in the southeast.

180

5.3. A missal of the highest standard: Seqv 33b add (Mi 107), s. xiii¹

5.3. A missal of the highest standard: Seqv 33b add (Mi 107), s. xiii¹ The remains of a truly remarkable missal are labelled Seqv 33b add (Mi 107) (Oslo, NRA Lat. fragm. 489 a.o.).

159

This was a very large missal (43 x 27 cm), written by a

very good scribe in a formal textualis, and richly decorated. The initials were gilded with gold leaf and embellished with “sprouts” touched with green and brown. Seventeen fragments from eleven leaves have survived, and one of the fragments is in the form of a bifolium. The bifolium was used to bind the book of the episcopal see (“stiftsboken”) of Oslo and Hamar from 1575, which is unusually early, as most accounts from before 1610 seem to have been discarded (cf. Pettersen 2003, 51). The other fragments are smaller, and seem to be examples of re-used bindings, since they are cut into smaller pieces and used to bind accounts for different places in Akershus in a later period, 1607-1639. That the oldest binding is by far the largest corresponds with the theory that earlier bindings were in fact to a larger degree bifolia (cf. Pettersen 2003). This missal is presumed to be produced in Norway and has been discussed by Karlsen (2005, 158-161). This chapter will take a closer look at some of the reasons for giving it a Norwegian origin, what models the scribe may have had, and who he (or they) was writing for.

5.3.1. A Norwegian scribe writing in English style The script is a textualis of highest formality, a textus praescissus (cf Derolez 2003, 76). In the praescissus the minims end flat on the baseline, sometimes rather wedgeshaped, or with a thin horizontal hairline. In Seqv 33b add the scribe has not been consistent, but has often turned the pen slightly to the right, failing to produce an even horizontal finish along the baseline seen in the best manuscripts of this type. The textus praescissus was used only for the most formal manuscripts, and was more popular in England than in the rest of Europe (Derolez 2003, 76). The general aspect of the script of this missal also displays English features, like the shape of the g and the ampersand. The ampersand occurs in combination with a crossed tironian note for “et”, indicating a date after 1200. Gothic features such as biting are not very

159

I am grateful for the use of the NRA database regarding Mi 107/Seqv 33b.

181

Part II: Analysis dominant and its use is rather random; round and straight d is used inconsistently, and round d can, for example, be used before e without biting. There are examples of cup-shaped suspension signs, used alongside straight suspension signs, which is rather peculiar. Since this manuscript was most likely written after 1200, the cup-shaped signs can be either a conservative trait in the scribe, or inspired from the exemplar. The letters are quite narrow, which would indicate that the missal was not written very early in the thirteenth century. Still, there is not a large degree of “breaking” in the arches of letters like m and n. It seems like a likely date for this manuscript would be the second quarter of the thirteenth century, which is not far from the mid thirteenth century date given by Karlsen (2005, 158 and the NRA database). As Karlsen has pointed out, the writing is above top line as opposed to below, a practice generally used in English manuscripts before the middle of the thirteenth century as first noted by N. R. Ker (1960b). To what extent this practice was followed in the different regions of Scandinavia is, to my knowledge, not carefully studied. Still, the breviary-missal Seqv 38 (Br-Mi 3), dated s. xiii med., appears to be written below top line, and may at least serve as an example that the change of practice was also to some extent followed in Scandinavia. But while Seqv 38 add still has double vertical lines framing the writing space, which is the earlier ruling pattern later replaced by a Fig. 26: Seqv 33b add (NRA Lat. fragm. 489, new number: Fr.628) The lower half of an initial I demonstrates the splendor of the missal. Notice the inconsistent textus praescissus and the use of small cap R in words like [lux]uRiose and vocaRi.

182

5.3. A missal of the highest standard: Seqv 33b add (Mi 107), s. xiii¹ single line (Derolez 2003, 37), Seqv 33b add seems to have a single line vertically framing the writing space, but a double line between the columns, which may be considered a transitional phase of the ruling practice. So why is this manuscript presumed to be produced in Norway and not in England? As pointed out by Karlsen (2005, 159), there is an extensive use of small caps

R

in

medial position, a feature which is considered to be an indication of Norwegian origin. Although small cap

R

does occur in English books, particularly pre-1200, it is

found mainly in nomina sacra and in initial or final position (Derolez 2003, 91). One may wonder if the textus praescissus in English manuscripts, even after 1200, invites to a more extensive use of small cap

R

than other script-types. Still, the use in Seqv

33b add seems too extensive for an English manuscript. There is one more element pointing to Norway or Scandinavia rather than England. The holes used to sew the original binding are still visible in the preserved bifolium. These are made as horizontal slits, not as holes, which would be the usual English procedure. Of course one could imagine that a book of this size would have been moved unbound and bound on arrival, as big wooden boards would make the book heavier and more troublesome to move. Although no feature is an absolute criterium for a Norwegian or Scandiavian provenance, three of the mentioned features point in this direction: It is likely that an English scribe would have mastered the textus praescissus better and been more consistent in the execution of this type of script, and the use of the small cap

R

seems to be too extensive for an English manuscript. In

addition there is the marginal slits from the original binding. I therefore conclude as Karlsen, that this manuscript is probably of Norwegian origin.

5.3.2. Possible models Which books would serve as models for this kind of manuscript? Several English manuscripts preserved among the medieval fragments could be of the kind used, although as mentioned below it has not been possible to identify the exact model for the contents. I would like to point to two examples – one missal and one psalter, both

183

Part II: Analysis datable to the first quarter of the thirteenth century. They are both written in textus praescissus and have the same basic type of initials.

160

As we see from Lat. fragm. 489-

11, the surviving initials seem all to have been gilded on a red base. While the gold has been rubbed of, the red base remains (see fig. 27).

Fig. 27: Seqv 33b (NRA Lat. fragm. 489-11, new number: Fr.614). The initial S, which was protected under the fold, still has its gold leaf more or less intact, while the more exposed D has only the red base left.

A missal provides the first example of one possible model for a scriptorium. It was presumably written in England in the first quarter of the thirteenth century, and three fragments remain in the Stavanger State Archive (Fragm. 1-3, by Lilli Gjerløw labelled Mi 35, along with a group of fragments in the NRA). There are two initials remaining, although these are not plated with gold, but are red and blue.

160

Karlsen also referred to the initials as being of an English type, and points to several examples of later and more simplified versions of the same type (Karlsen 2005, 160). I will, however, point to initials of the same degree of elaboration, which could have been in Norway at the time Seqv 33b add was made.

184

5.3. A missal of the highest standard: Seqv 33b add (Mi 107), s. xiii¹

Fig. 28: Stavanger, statsarkivet, fragm. no 1-3 (Mi 35). Remains of an English missal antedating the Seqv 33b, presumably used in Norway already in the Middle Ages.

There are several similarities between this English missal and Seqv 33b add: They were both books in folio format. The scribe used both the ampersand and the crossed tironian note for “et”. There is also the same inconsistent use of biting. Here the text is also written above top line, but the writing space is framed by double vertical lines, with a narrower space between them than the horizontal text lines, a trait of the transitional period of the twelfth and thirteenth century (Derolez 2003, 38). The “sprout”like decorations are here smaller and rounder, but still basically the same type as that of the missal above. The second example is a psalter (Oslo, NRA Lat. fragm. 103), along with other psalters of the same basic type. The collection at NRA holds a number of high quality psalters of the late twelfth or early thirteenth century with the same type of goldplated initials with “sprout”-like decorations.161 161

For example NRA Lat. fragm. 96 and 97 (same psalter); 103; 107; 108.

185

Part II: Analysis

English liturgical books with the same basic type of decoration are also quite common in the Swedish material.

162

Although such books would have been

purchased for the wealthiest Scandinavian patrons only, there would have been a few around to serve as inspiration and direct models for local scriptoria of a high enough standard to be able to produce manuscripts like the missal Seqv 33b. If this missal was produced in Norway, it would have been a top quality product of its scriptorium. While a psalter of such degree of decoration like the ones in the NRA could have been written for a wealthy lay-person, a missal of this type could have been written to be used in a cathedral, a large wealthy town-church or a royal chapel.

5.3.3. Sequences not secundum ordinem So far the contents still have not been treated. No sequences are written out in full, but five sequences are entered with incipits. This may either mean that they were written out in full at an earlier point, or that the missal was intended to be used in combination with a separate sequentiary, either at the end of the Fig. 29: Oslo, NRA Lat. fragm. 103. Thirteenth century psalter.

book, or as a separate volume. It would have been convenient if the missal turned out to follow the Nidaros ordinal. It does not.

One of the peculiar traits of the contents is the use of the sequence Pro nobis ora for (the octave of?) St. Stephanus. This sequence is not in the Nidaros ordinal, and has not been found in other Norwegian or Icelandic manuscripts with sequences. It is also not found in the Swedish material (cf. Björkvall 2006, 58) and is not common in either the Anglo-French or the German sequence repertories. A sequence Pro nobis ora (AH 37, no. 321) to the melody Inviolata is registered for St. Vincentius in two 162

See for instance SRA, Fr 7134 (http://www.ra.se/ra/MPO-exempel.html), or Abukhanfusa 1993, 76 (Fig. 62).

186

5.3. A missal of the highest standard: Seqv 33b add (Mi 107), s. xiii¹ fourteenth century missals from Metz and a printed missal from Toul (both bishoprics of Lorraine), but whether this is the same sequence, is not certain. The fragment also contains a sequence-like rendering of the text Factum est autem. I have been unable to link these sequences to any particular use, but the contents may lead us to the use at a later date. It would be interesting, as the registration of the NRA fragments continues, if there were more “local” books of this quality.

187

Part II: Analysis

5.4. A manuscript with unknown sequences: Seqv 25, s. xiii¹ The sequentiary Seqv 25 (Oslo, NRA Lat. fragm. 532) consists of two small fragments used to bind an account from Fredrikstad for the year 1611. This sequentiary is interesting for several reasons. Its scribe was quite good, and wrote an English-type script from the first half of the thirteenth century. The scribe could either be English, or a skilled Norwegian trained to write in English style. The one remaining large initial is monochrome green, and the smaller initials are red, green and yellow (or light ochre), which is a rather old-fashioned palette for the thirteenth century, when one – at least for an English manuscript - would expect the initials to be red and blue.

5.4.1. A sequence for St. Andrew? The unusual thing about the manuscript is its contents. In addition to the sequence Pangat nostra contio (AH 9, no. 390), in the Nidaros ordinal used for confessor bishops, it contains two otherwise unknown or unidentified sequences. The first sequence contains only parts of the two verses of the final strophe: x. y.

•••• •••• | • • • • intuere | dator indulgentie. Et ex• • • • • •| • • • • • • • le | locum et clementie.

(• marks the assumed number of missing syllables)

Lilli Gjerløw suggests that these might be the last words of the otherwise untraced sequence for St. Andrew (30 Nov.), Precluis ecclesia (Gjerløw 1968, 433).163 What this hypothesis was based on is not quite clear. The melody of the final verses seems to resemble that of “De profundis tenebrarum” for St. Augustine, although the emphasis on the melodic movement seems to have shifted.164

163

For the other untraced sequences, see ibid, and Kruckenberg 2006a, 18. While De profundis starts the final section of the final verse on the f, moving up to a, c and d, touching the e, the Unknown ends the previous section on the f, and starts the final section of the final verse by moving up to a, c, d, without touching the e. For the melody of De profundis tenebrarum see either Eggen 1968 I, 241 (who supplied the melody from Moberg), or Moberg 1927 II, no. 19.

164

188

5.4. A manuscript with unknown sequences: Seqv 25, s. xiii¹

Fig. 30: NRA Lat. fragm. 532-2. This fragment contains two unknown sequences, one with the explicit "locum et clementie", the other with the incipit "Quare fremuerunt gentes".

5.4.2. Quare fremuerunt for the Holy Blood? Fortunately there is more to go on for the other “unknown” sequence, including the incipit: Quare fremuerunt gentes. The remaining parts of the text are as follows, with lacunae: 1a. 1b.

Quare fremuerunt gentes | c• • • • • • • • Fides constans agoniste | • • • • • • • •

| • • • • fidei | • • • • • pii

2a. 2b.

Consilium malignantium ob[sedit]…[do]minum regnantem in eternum. A... [lacuna]

3a. 3b.

•••• •••• Tibi testis o lau• •

4a. 4b.

Promat enim di• • • • | • • • • • • • • Per ifrastes pal• • • |•••• ••••

5a. 5b.

Iubilemus • • • • Sanguis • • • • • •

6a.

[lacuna]

|•••• •••• | sed divina gratia. |•••• •••• | • •165 est audatia. | • us est eulogio. | absit a consortio.

|•••• ••••l | in novis miraculis. |• • • • • us preco | predicans in populis.

(• marks the assumed number of syllables, while ... means that the number of syllables is unknown.)

165

Possibly “fracta” (?), cf. Augustinus Hipponensis, Enarrationes in psalmos. In psalmum 98 enarratio, sermo ad plebem (PL 37). Source: Patrologia Latina. The full text database: http://pld.chadwyck.co.uk/

189

Part II: Analysis The first strophe and the incipit are the first words from Psalm 2: “Quare fremuerunt gentes et populi meditati sunt inania, adstiterunt reges terrae et principes convenerunt in unum adversus Dominum et adversus christum eius.” (Why do the nations conspire, and the peoples plot in vain? The kings of the earth set themselves, and the rulers take counsel together, against the Lord and his anointed) The other half-strophe is dedicated to the unwavering faith of the “agonista”, the combatant or warrior. The second strophe recalls Psalm 21, 17: “Quoniam circumdederunt me canes multi, concilium malignantium obsedit me, foderunt manus meas et pedes meos.” (For dogs are all around me; a company of evildoers encircles me. My hands and feet have shriveled). The rhythm is not the same in the first and second strophe or versepair. While the first strophe is trochaic, the second does not seem to be adhering to any particular verse-form. The second verse in strophe three, beginning with “Tibi testis o lau…”, may have held the clue to whom this sequence was written. Lilli Gjerløw points out in her unpublished catalogue in Oslo, NRA, that there are three options: Olau[e] O lau[renti] O lan[berte] The Lanbertus in question would be Lambertus of Maastricht (d. 705), martyr and bishop, whose feastday is 17 September. He was speared from behind in his house, possibly for criticising Pipin of Herstal for leaving his wife for his mistress. Lambertus was particularly revered in the Low Countries, Belgium and Germany. The feast in natalicio Lanberti martyris ac pontificis is entered in the Nidaros ordinal with nine lessons and the sequence Celi enarrant (Gjerløw 1968, 395-6). Still, the letter does look more like the letter u than the letter n. This alternative does not seem very likely.

190

5.4. A manuscript with unknown sequences: Seqv 25, s. xiii¹ If the text reads olau[e] we would be looking at an unknown sequence for St. Olav, which would truly be remarkable. However, “olaue” would not fit well with the trochaic rhytm which seems to govern most of the strophes in the sequence, with the exception of the second strophe. In addition there seems to be a space between “o” and “lau”. The name Laurentius in the vocative case, Laurenti, would fit well with the verseform (“Tíbi téstis ó laurénti”). This is a possible alternative. There is at least one fourth alternative which is that “lau-“ is not the beginning of a saint’s name at all, but an epithet like “o laudabilis” or, which would work better with the trochaic verseform, “o laudande” or something similar. If this is the case, the sequence could be for practically anyone, although probably still a martyr because of the general references to struggle and the blood in the fifth strophe. It could also be an adjustable sequence for an unspecified martyr, although the use of an unusual word like “periphrastes” could be a reference to something specific. The fourth strophe does not seem to hold the key to the sequence’s assignment. The verb “promo” (bring forth, let be heard) is used in other sequences, like Prome casta contio or the strophe “novum promat canticum” in the twelfth century Parisian sequence Splendor patris et figura (cf. Fassler 1993, 158). Whatever is uttered seems to be done so with praise (eulogio). “Per ifrastes”, probably for “periphrastes”, refers to someone rephrasing the words of others and it could be used in both a positive and negative sense. Whatever word “pal-“ is part of (“palatinus”? - of the palace, imperial, or “palliatus”? - wearing the pallium), it forms a nice alliteration with “promat” and “perifrastes”. The fifth strophe, and the final in this fragment, encourages us to rejoice in the new miracles, and the last verse is about a herald preaching to the people. In the most recent book on the Nidaros sequences Quare fremuerunt is suggested as one of the sequences for the mass of the Holy Blood (14 Sept) (Kruckenberg 2006, 35). The feast of the Holy Blood was a local feast in the diocese of Nidaros proper celebrating the arrival of a drop of the blood of Christ (presumably kept in a ring) in 1165.166 The 166

The latest research into this relic the liturgy of its feast is available in The Nidaros Office of the Holy Blood (Attinger and Haug 2004).

191

Part II: Analysis presence of the word “sanguis” in verse 5b could certainly be an argument for this hypothesis. In that case Quare fremuerunt was a sequence written for use in Nidaros proper. It would therefore have been convenient if the region indicated on the account had belonged to Trondheims len. This is unfortunately not the case, since Fredrikstad (which the tax-account was written for) is far from Trondheim. Still manuscripts moved around both in medieval times and post-Reformation, so it is certainly not impossible. The common topic of the psalm-quotations seems to be connected to some kind of struggle with opponents. In the first and second strophe the sequence tells of conspiring peoples, the unwavering faith of the warrior, a company of evildoers and the Lord ruling for ever. The blood in the fifth strophe could refer to that of a martyr, or possibly to the blood of Christ himself, as recently suggested. It is unfortunate that the sequence appearing just before Quare fremuerunt is unidentified. A known sequence in this position could have helped us to place Quare fremuerunt in some kind of ecclesiastical context. It is also unfortunate that we can draw no firm conclusion regarding the origin of the sequentiary. If the sequentiary is English, there is no reason to presume that the unknown sequences it contains are connected to Norway. If, on the other hand, we had any convincing arguments that the sequentiary was written in Norway, these sequences could be evidence of local sequence production and added to the sequences for local saints already identified.167

167

Lux illuxit (AH 42, no. 302), Predicasti dei care and Postquam calix babylonis for St. Olav, Lux illuxit for St. Hallvard, and possibly Laudes debitas deo nostro (AH 54, no. 62) for St. Lucia/St. Agatha.

192

5.5. Three local sequentiaries – and one roll: Seqv 1 et alia, s. xiii¹

5.5. Three local sequentiaries – and one roll: Seqv 1 et alia, s. xiii¹ By far the most extensive sequentiary (and troper) in the NRA is labelled Seqv 1 (Oslo, NRA Lat. fragm. 418), for it comprises 46 fragments from 21 leaves, none of which complete. The sequentiary in its present state contains 29 sequences, but the original number was probably at least 50, based on the number of missing feasts (Ommundsen 2006, 150). Two other sequentiaries are comparable to Seqv 1 in either contents or form, both with a secondary provenance in eastern Norway, namely Seqv 2 (Oslo, NRA Lat. fragm. 336) and Seqv 15 (Oslo, NRA Lat. fragm. 419). In this chapter it will be discussed how these three local thirteenth century sequentiaries relate to the Nidaros ordinal.

5.5.1. Seqv 1 Seqv 1 is a “classical” troper-sequentiary in the Anglo-French tradition. The sequentiary begins with the sequence for the first Sunday of Advent, Salus eterna (AH 7, no. 4/53, no. 1). Seqv 1 is the oldest source to contain Lux illuxit letabunda for St. Olav (29 July). The presence of St. Olav’s sequence is the main argument for assigning Seqv 1 a Norwegian origin, since the script is of a type and quality which could just as easily have been written in England or France. The date of the sequentiary is probably some time in the first half of the thirteenth century, most likely the second quarter. This means that the sequentiary was probably copied after the introduction of the Nidaros ordinal, but it does not comply more with the Nidaros ordinal than any other Anglo-French sequentiary would, with the exception of the St. Olav sequence. The secondary provenance of the sequentiary was most likely eastern Norway, and, most unusual, it is even possibly to suggest a specific church, namely Hedenstad church south of Kongsberg (for Kongsberg, see map in chapter 1.1.4.). In a fragment discovered by Odd Sandaaker as late as 1978 there was a marginal comment, saying “Dominica ante festum diuj olaj regis adfui ecclesie Hedenstad Ego sane mirabilia vidj opera rusticorum” (“On Sunday before the feast day of St. Olav the king [29

193

Part II: Analysis July] I was present in the church of Hedenstad. I saw the truly amazing rustic works”) (quoted in Gjerløw 1988). Lilli Gjerløw drew the obvious conclusion that the sequentiary was in Hedenstad church when this was written by a visitor (ibid.). She also pointed out that the fiefs Numedal, Eiker and Brunla, from which the accounts bound in Seqv 1 fragments came, in the period in question were held by the same feudal lord, the Danish nobleman Ove Gjedde (1594-1660). Gjedde resided in Kongsberg, and it was probably here the sequentiary was dismantled (Ommundsen 2006b, 144). The sequences in Seqv 1 are presented in the table below. Feastdays for which there are no preserved fragments, but which were probably in the complete Seqv 1, are presented in the left column. For a further discussion and comparison with other Anglo-French sequentiaries, see Ommundsen 2006b. Table 8: The contents of Seqv 1 (Lat. fragm. 418) Feastday (temporale) Dom. i adventus Dom. ii adventus Dom. iii adventus Dom. iv adventus Natale dni ad primam missam Ad secundam missam Ad tertiam missam Stephani (26 Dec.) Iohannis (27 Dec.) Innocentium (28 Dec) Thome Cantuariensis (29 Dec) Sexto die nat. (si dominica) Silvestri pape (31 Dec) In circumcisione dni (1 Jan) In epiphaniam (6 Jan) Dominica resurrectionis Feria ii-viii post resurrect. Inventio s. crucis (3 May) In ascensione In die pentecosten Feria ii post pent. Feria iii post pent. (?) Feria iv post pent. Dominica post pent. Dom. ii post pent. Dom. Iohannis bapt. in crast (25 June) In divisione apostolor. (15 July)

Seqv 1

Ordo Nidrosiensis Ecclesiae

Salus eterna Regnantem sempiterna Qui regis sceptra Iubilemus omnes una Nato canunt omnia lacuna

x x Gaudia mundo x x

Magnus deus Iohannes Iesu Christi lacuna

x Virgo mater

Laudes crucis Rex omnipotens lacuna

x x

Veni sancte spiritus Laudes deo devotas Benedicta sit beata Quicumque vult salvus esse Voce iubilantes magna Sancti baptiste lacuna

Eia musa Almiphona iam gaudia x O alma trinitas x x

194

5.5. Three local sequentiaries – and one roll: Seqv 1 et alia, s. xiii¹ Marie Magdalene (22 July) ? Olavi regis et martyris (29 July) Laurentii in die (11 Aug)

Mane prima sabbati possible lacuna Lux illuxit letabunda Laurenti David

De assumpt. sce Marie (15 Aug) Feria ii post assumpt. Feria iii-viii Nativitas Marie (8 Sept) ? Michaelis archangeli (29 Sept) Omnium sanctorum (1 Nov) In natali Martini (11 Nov) Andree ap. in die (30 Nov) Nicholai (6 Dec) unius apostoli ? ? unius virginis

Aurea virga Stella maris o Maria lacuna Alle celeste nec non possible lacuna Ad celebres rex celice Christo inclita candida Sacerdotem Christi Sacrosancta hodierna Congaudentes exultemus Clare sanctorum senatus possible lacuna Adest nobis dies alma Virginis venerande

x vel Laus tibi Christe x Martyris eximii Laurenti David in crast. x Congaudent angelorum x x x x x -

5.5.2. Seqv 2 The other two sequentiaries also seem to be of an Anglo-French tradition and not following the Nidaros ordinal. In the case of Seqv 2 it might be that it precedes the release of the ordinal, as its date is probably in the first quarter of the thirteenth century. The script is of an English or French type, as can be seen in the shapes of the g’s. The scribe, on the other hand, was most likely Norwegian. The argument for this is not only the stiffness of the hand, but also the occurrence of an æ-ligature in Celsa pueri, which seems to be a misreading of an ampersand for “et” in the word valet. That it is an æ-ligature in this case, and not an ampersand, is illustrated by the ampersand following immediately after (

).168 An æ-ligature indicates

influence from a Scandinavian vernacular, as this phenomenon does not seem to be common in England. The early date manifests itself both in the Protogothic nature of the script, and in the pigments used in the larger and smaller initials: in addition to red and blue, yellow is also one of the main colours both in the larger and smaller initials (although not in the facsimile below). The hand is in fact not so different from the hand responsible for the property list in the Munkeliv Gospelbook (Copenhagen, GKS 1347 4°) in terms of the lettershapes and quality (cf. Hødnebø 1960, Pl. 1)

168

See Eggen 1968, Pl. 46.

195

Part II: Analysis

Fig. 31: These lines from Seqv 2 (Oslo, NRA Lat. fragm. 336) show the hand of the scribe, who probably wrote the sequentiary in the first quarter of the thirteenth century.

Far less remains from Seqv 2 than Seqv 1, in fact only six sequences for the period between the third Sunday of Advent and the Circumcision (1 Jan). Table 9 below demonstrates how Seqv 1 and Seqv 2 relate to the Nidaros ordinal and other English and French sequentiaries of approximately the same time. Table 9: Seqv 1 and Seqv 2 and selected English and French sequentiaries in relation to the Nidaros ordinal for Advent, Christmas and the Circumcision Feastday (temporale) Dominica I Adventus Dominica II Adventus Dominica III Adventus Dominica IV Adventus In nativitate domini, prima m. In nativitate domini, secunda missa In nativitate domini, tertia m. De st Stephano (26 Dec) De st Iohanne ap.ev. (27 Dec) De st Innocentibus (28 Dec) De st Thoma martyre (29 Dec) Sexto die natalis Christi (30 Dec) De st Silvestro (31 Dec) In circumcisione domini (1 Jan)

Sequences, Nidaros ordinal Salus eterna Regnantem sempiterna Gaudia mundo

Seqv 1 x

Seqv 2 Lacuna

x

Chichest. Ox 148 x

Dublin Cdg 710 x

x

x

Qui regis sceptra x

Qui regis sceptra x

x

x

x

x

x

Lacuna

Lacuna

Lux fulget hodierna

Letabundus exultet

Letabundus exultet

x

Magnus deus

x

Christi hodierna Lacuna

x

Celeste organum x

Christi hodierna x

Virgo mater

Iohannes Iesu Chr Lacuna

Laus devota x

Iohannis Iesu Chr x

Gratulemur ad f.

Circa fines

-

Sollempne canticum Christi hodierna -

x

x

Iubilemus omnes una Nato canunt omnia Celeste organum vel Lux fulget hodierna Celica resonant

Celsa pueri Spe mercedis

x Lacuna

Lux fulget hodierna Panga nostra contio Eia recolamus

Qui regis sceptra x

-

x

196

Qui regis sceptra x

St. Denis Pa 1107 Gaudia mundo x Qui regis screptra x

5.5. Three local sequentiaries – and one roll: Seqv 1 et alia, s. xiii¹ The table shows that neither Seqv 1 nor 2 follow the Nidaros ordinal for the third Sunday of Advent, but have Qui regis sceptra, in accordance with the Anglo-French tradition. For the third mass on Christmas day, Seqv 2 also has a different sequence than the ordinal. The other sequences remaining in the sequentiary are common ones. Those sequences which display a wider variation, like those for St. John the evangelist (27 Dec) and St. Thomas of Canterbury (29 Dec) are unfortunately missing.

5.5.3. Seqv 15 The third manuscript resembles Seqv 1 in lay-out and style, although it is a bit larger. It was probably written at around the same time, i.e. the second quarter of the thirteenth century. The scribe of Seqv 15, who was probably Norwegian, is also quite good, but does not show quite the same confidence. As one of the remaining leaves from this manuscript does not contain sequences, but elements from a gradual, the sequentiary was probably the final section of the gradual. The provenances of the accounts it was used to bind is Fredrikstad and Tunsberg, on either side of the Oslofiord, and I have not been able to establish a logical connection, as Tunsberg was under the nobleman Gunde Lange (c. 1570-1645), but Fredrikstad was not. Akershus could be a likely origin for the bindings in this case. Although only five sequences remain from Seqv 15, one of them is Stella maris (presumably to be sung within the octave of the Assumption of Mary 22 Aug.), which is not prescribed in the Nidaros ordinal (cf. table above with the contents of Seqv 1). Seqv 15 also contains other sequence for the octave of the Assumption, like Ave mundi spes and Alle celeste nec non, but unfortunately Seqv 1 has a lacuna here which prevents comparison. Seqv 15 is the third local thirteenth century sequentiary in this chapter which does not comply with the Nidaros ordinal. While Seqv 2 very well could have been written before the introduction of the ordinal, Seqv 1 and 15 were probably copied after the ordinal was spread. So why did scribes copy classical Anglo-French sequentiaries even after the Nidaros ordinal? A possible explanation could be that

197

Part II: Analysis when a new sequentiary was needed, the exemplar governed the new copy – not merely because it was easier copying a book without altering the contents, but also because an existing book would hold a certain authority. An exemplar – and certainly if this was an exemplar which to some degree had governed the local practice - would represent a certain counter-force to the ordinal. While the contents of a central Mass book like the missal would probably be given more attention with regard to the Nidaros rite, a sequentiary was usually a type of liturgical book with many variations, some small, some large, and some having an Anglo-French repertory, some a German one. To what extent local churches performed sequences prescribed by Nidaros, or were allowed (or took) some freedom in their sequence repertory, governed by local tradition, is not known. Perhaps some cantors would prefer to celebrate a church feast with a song they knew (and perhaps cherished) than an officially prescribed song which may have been unknown to them.

5.5.4. A roll A reminder that sequences were not only spread through cantors or books is a parchment roll found under the floorboards of the stavechurch in Lom during excavations in the 1970’s (Museum of Cultural History, University of Oslo, C 34738/52). Its scribe looks Norwegian, writing in the middle of the thirteenth century. The text, a sequence for St. Thomas of Canterbury, Aquas plenas (AH 55, no. 326), not prescribed in the Nidaros ordinal, was edited by Lilli Gjerløw (1978). The sequence was equipped with musical notation, but the larger and smaller initials were not filled in. Because of the traditional flexibility of sequence repertories, and the likelihood of a large variety in the use of sequences in local pre-ordinal rites, a scribe in a larger centre would probably have a multitude of sequentiaries to choose from, with varying contents. If a scribe chose one book to follow faithfully (as suggested above), some of the sequences “missing” according to the ordinal could probably be found in other sequentiaries. Or if a new sequence needed to be introduced, the most

198

5.5. Three local sequentiaries – and one roll: Seqv 1 et alia, s. xiii¹ convenient thing would be to acquire it on a parchment roll, and, if desired, the sequence could later be copied into a new volume, like Lux illuxit in Seqv 1. The use of rolls for sequences is attested in one of the first (and most important) written sources regarding the sources, namely Notker of Sankt Gallen’s proemium to the Liber hymnorum. After making his first sequences, he went to his teacher and showed them to him: Quos versiculos cum magistro meo Marcello praesentarem, ille gaudio repletus in rotulas eos congessit; et pueris cantandos aliis alios insinuavit (Steinen 1948, I, 10). When I presented these small verses to my teacher Marcellus, he was filled with joy and had them copied on to rolls, and assigned them to the boys to be sung, some to some, others to others.

The “in rotulas eos congessit” has formerly been translated “had them copied as a group on a roll” (Crocker 1977, 1) and “sammelte er sie ... auf Einzelblättern” (Steinen 1948, I, 11). I think the plural is significant, as it would make more sense to be able to hand the rolls round than to have a group of sequences on one roll. I would also prefer to translate “rotulas” with “rolls” rather than “Einzelblättern”, although single leaves could probably be used in the same way. While the text quoted above testifies to rolls being used for rehearsals, medieval book illustrations show that rolls were also commonly used in performance. One example in a manuscript is the British Library, Arundel MS 83, f. 63v (reproduced in Bell 2001, 47 and cover), where three singers have their attention directed at a roll with text and musical notation on the music stand. One of the singers also seems to be holding a rolled-up roll in his hand. Parchment rolls are not included in the estimate of medieval books in Part I, and how many rolls may have existed is not known. An unprotected roll would be even more vulnerable than a bound book.

199

Part II: Analysis In this chapter I have introduced “the authority of the exemplar” as a possible explanation as to why Anglo-French sequentiaries seem to have continued to be copied in the thirteenth century, even after the introduction of the Nidaros ordinal. I want to emphasise that not all scribes would have followed their exemplars as faithfully – on the contrary, some would probably have found it useful to adapt their own book to correspond with the ordinal so far as they were able to, perhaps using two or more exemplars to supply the prescribed sequences. Whether or not a “Sequentiarium Nidrosiensis Ecclesie” were ever issued will be briefly discussed in the next chapter. To what extent local churches followed the Nidaros ordinal in practice is, at least at present, not possible to say. What seems clear is that although thirteenth century sequentiaries existed which were not in compliance with the Nidaros ordinal, this should not automatically be taken as evidence of “anarchy” in local churches. Sequences of the ordinal missing from one particular sequentiary could be supplied through other books or through parchment rolls.

200

5.6. Manuscripts secundum ordinem: Seqv 18 and 38 add, s. xiii med

5.6. Manuscripts secundum ordinem: Seqv 18 and 38 add, s. xiii med In the last chapter we saw evidence to suggest that in the thirteenth century sequentiaries were still being produced in accordance with older exemplars, with no apparent effort to make the books comply with the Nidaros ordinal, perhaps with the exception of the inclusion of St. Olav’s sequence Lux illuxit. This chapter will take a closer look at two “broader” liturgical books with sequences which remain from the middle of the thirteenth century, both with a content and a form which point in the direction of a Norwegian origin and a relationship with the Nidaros ordinal.

5.6.1. A missal and a breviary-missal A missal and a breviary-missal, both written around or just after the middle of the thirteenth century, are among the manuscripts with sequences entered among the other liturgical elements the celebration of Mass. The missal is labelled Seqv 18 (Mi 106), and the 14 fragments from 9 leaves were used to bind accounts from Trondheim in the period 1624-27 (Oslo, NRA Lat. fragm. 881, 882, 932; New numbers Fr.600-613). The breviary-missal Seqv 38 add (Br-Mi 3) was identified by Lilli Gjerløw, and seven fragments from four leaves were found under four different fragment numbers, three in the NRA, used to bind accounts from Stavanger len 1651-54 (Oslo, NRA Lat. fragm. 668, Sandakers kat. LR pk. 288 F. Jæren og Dalane 1651-52 and pk. 34 C. Ryfylke 1651), and one fragment from the same books was found in Stavanger State Archive (SAS, Fragm. 13). The two books have several features in common: They were both written in Norway at more or less the same time, around the middle of the thirteenth century. They were of approximately the same size (c. 20 x 15 cm), although the leaves are now cropped, in a format which would be handy for a travelling priest bringing with him his own missal or breviary-missal. The scribes in both manuscripts write in a quite informal way, and basically write an “English” style script (as we can tell from the ductus of the g, which in both mss is also influenced by documentary script, and the shape of the punctus elevatus). The initials are modest. The missal uses the pigments

201

Part II: Analysis red, blue and green, while the breviary-missal uses red and green, with red penflourishing on the larger green initials and vice versa. Another thing these books have in common is that they seem to be written in general accordance with the Nidaros ordinal. That several fragments from the missal Seqv 18 over a three-year period were used to bind fragments from Trondheim, points in the direction of the Trondheim area for its medieval secondary provenance. The provenance for the breviary-missal, however, is the Stavanger area, as is supported by the fact that one fragment from the breviarymissal was never sent to Copenhagen and still remain in Stavanger State Archives. Whether or not their origin is somewhere close to their provenance is not easy to say, but it does seem likely. The scribe of the missal writes with a certain “sharpness” and with a relatively thin nib. Blue and red seem to be the dominant colours, but green pigment is also used. In the breviary-missal there is no sign of blue colour, and the use of red and green, along with the round-shaped letters written with a thick nib, may connect it to other books of a presumed origin in the western parts of Norway, probably Bergen. The red and green of the books of western Norway may be said to stand out at a time when the common practice was to use red and blue for initials. Both the missal and breviary-missal may serve as examples of the kind of books which were presumably produced in large numbers after the introduction of the Nidaros ordinal. They may also be examples of regional book production for Bergen and Nidaros respectively, although to make any conclusions at this point would be premature. First we need a better overview of the rest of the material, to find relevant comparable fragments. The missal Seqv 18 seems to have been written at greater speed than the breviarymissal, as shown through the execution of the a’s, but with confidence. The slashes over the i’s are very distinct, and give, along with the marked serifs on the descenders of the p’s and q’s, a certain dynamic to the page. At the same time there are some old-fashioned features, like the mid line punctus, and some suspension

202

5.6. Manuscripts secundum ordinem: Seqv 18 and 38 add, s. xiii med signs looking almost cup-shaped or wavy. The ruling is so weak that it is difficult to see if the scribe wrote below or above top line. Seqv 38 add, on the other hand, upon close inspection does appear to be written below top line. This is an important example, showing that the phenomenon described by Ker (Ker 1960b) reached Scandinavia early. To write below top line is a modern feature, but here it occurs together with features which seem old-fashioned: there is very little “breaking” in the letters, and there is hardly any biting. The d’s are mainly straight, and although there are some examples of round d in fusions with e and o, particularly in the word “domino”, round d actually seems to be favoured in final position. Another feature, which is quite unusual, is the use of small cap

H

in

words like Habent and Hostias, i.e. not nomina sacra. Of the two manuscripts, Seqv 38 add would probably be labelled a Norwegian or Scandiavian product based on the script alone, even before taking the contents into consideration, because of the unusual mixture of conservative and late elements, and the use of the small cap

H.

The missal does not stand out to such a degree, apart from that it is written informally and quickly, to a larger degree resembling documentary script. In the case of Seqv 18 it is placed in Norway through the presence of St. Olav’s sequence Lux illuxit and the correspondence with the Nidaros ordinal. Coincidentally, both manuscripts contain the liturgy for feasts in summer and early autumn, although they do not “overlap”. In addition to other liturgical elements, the missal Seqv 18 contains six sequences: Petre summe (SS. Peter and Paul, 29 June), Sollempnitas sancti Pauli169 (St. Paul, 30 June) [lac.], Lux illuxit (St. Olav, 29 July) [lac.], Martiris eximii (St. Lawrence, 10 Aug) [lac.], Grates, honos, hierarchia (in exaltatione S. Crucis, 14 Sept),170 and Virginis marie laudes (BMV, incipit only). The rest of the contents also correspond with the ordinal. The lacunae between the surviving fragments include at least two feasts which would have been interesting to have evidence of, namely St. Swithun (of Winchester, patron saint of Stavanger, celebrated 2 July) and the Saints of Selja (= Sunniva and her followers, patron saints of Bergen, 169

Identified by Eggen as Laudes Christo canamus (Eggen 1968, 150), but corrected by Lilli Gjerløw to Sollempnitas sancti Pauli, because of the context. 170 Grates honos has until now been unidentified in this manuscript.

203

Part II: Analysis feast day 8 July). The liturgy, including sequences, for these saints has not yet been discovered in any Norwegian manuscript fragment.

171

Seqv 18 does not correspond

completely with the Nidaros ordinal, as it goes directly from the feast of St. Lawrence (10 Aug) to the Assumption of St. Mary (15 Aug), without mentioning the Fig. 32: Seqv 18/Mi 106 (Oslo, NRA Lat. fragm. 882-5, -6), with the incipit for Virgini marie laudes.

171

The liturgy prescribed for these saints in the Nidaros ordinal is known (and transmitted in the printed Missale Nidrosiense). For St. Swithun the ordinal gives the alternatives Psallat ecclesia mater decora and Pangat nostra contio. The sequence prescribed for Sunniva and the other saints of Selja is Ecce pulchra from the commons. It would have been interesting to see if this sequence was still used in the mid thirteenth century, or, whether it had been replaced by a sequence proper for the saints.

204

5.6. Manuscripts secundum ordinem: Seqv 18 and 38 add, s. xiii med celebrations of St. Tiburtius (11 Aug) and St. Ypolitus (13 Aug) in between. It also prescribes the sequence Virgini marie laudes for a mass for St. Mary, a sequence which is not in the Nidaros ordinal.

5.6.2. Grates honos hierarchia One sequence is present in Seqv 18/Mi 106 only in partial words, and it has not previously been identified. The remaining parts of the sequence are as follows: [...]ntido-

sor[...]

[...]alissimum.

Nos [...]

[...]ia cle-

ta [...]

[...]on. (?)

du[...]

[...]ulce

nte[...]

[...]rum

do[...]

[...]ucis

post[...]

The presence of words like “[d]ulce” and “[cr]ucis” made it reasonable to look for the text among the sequences prescribed in the Nidaros ordinal for the celebration of the cross, namely Laudes crucis or Grates honos hierarchia. The words or partial words in our two fragments (underlined in the strophes below) corresponded with the last of the two mentioned sequences for the cross, namely Grates honos hierarchia (AH 50, no. 239), prescribed in the Nidaros ordinal for the Exaltation of the Cross (14 Sept). The remaining pieces in Seqv 18 belong in the verses 5b, 6ab, 7ba and 8a (see below). The relatively rare Grates honos is one of the sequences of the Nidaros ordinal likely to have arrived in Nidaros as part of a German repertory, as it is otherwise not known outside southern Germany (Bower 2006a, 127). More specifically, it seems to be a sequence which can be connected to the Hirsau use, and it may have reached Nidaros through Lund. According to Lori Kruckenberg, who has studied the Hirsau sequence repertory (cf. Kruckeberg 1999), most of the sources containing Grates honos

205

Part II: Analysis have some connection to Hirsau.172 Until now only one Icelandic fragment with this sequence has been identified in Nidaros. The text below is taken from the most recent edition (Bower 2006b, 283-287). The verses of strophe 7 appear in the order they do in the fragment, which is the opposite of the edited text, both in Analecta hymnica and in Bower 2006b. Finally, the verses are not as frequently separated into lines as in the edited texts. Grates honos hierarchia, v. 5-8: 5b. Et contra mortis potyrium vitae propines antidotum. 6a. Tu totus desiderium, boni totius genus generalissimum, gaudimonium tu quam verissime yperbolicum, solaque tu sotheria, clemens tui nos intima pasce theorica. 6b. Theu panta eleymon aphesis benignula tu ton amartyon, sanctimonium dulce, iocundule tu deliciae portus quietis unice, archos patrum et optimas, eleyson ymas. 7b. Et tu solus qui fortiter crucis torcular tristeque prelum - vir de gentibus nullus tecum – idem ipse botrus elegans Cypri rubicundulus calcasti, bibens nobiscum potum te, nobis ipsum, tui fer patris in regno. 7a. Fac, nos calix inebriet perquam optimus sobrietatis, spiritalium dulcedinis, aeternorum mirae dilectionis, sophiaeque salutaris, quo vitis Sorech palmites fructus plures ferre queamus laetantes. 8a. Nos ut immolantes tui sanguinis sacri tibi rubentia musta cotidie mundicorditer et intime nudam crucem nuduli baiolemus carne et noy. 8b. Teque, dux, sequamur sponte voluntaria - non abre ut Simon et in angaria mundo re vera moriendo tibique nos, non nobis, vera vita, posthac vivendo.

172

I am most grateful to Lori Kruckenberg for correspondence regarding the Hirsau connection.

206

5.6. Manuscripts secundum ordinem: Seqv 18 and 38 add, s. xiii med The syllable sor, the first word on the verso-side of the leaf (see fig.), belongs to the word Sorech in verse 7a. The proximity of Sorech to the word Nos, the first word in verse 8a, shows that the verse which is most commonly labelled 7a came in the place of 7b in this fragment. This verseorder in strophe 7 is otherwise registered in AH (Dreves 1907, 311), and in Bower 2006b, 286. The other manuscripts displaying the verses of strophe 7 as here was labelled and identified by Bower as La (London, British library, Add 11669, gradual, sequentiary and sacramentary written in Augsburg c. 1150), M1 (Munich, Bayerische Staatsbibliothek, clm 13845, sequentiary and troper written in Sankt Emmeram in early twelfth century) Sf (Stockholm, Royal Library, Pap.

18

4°,

a

manuscript

of

Icelandic

provenance) (cf. Bower 2006, 279-80). In other words, the two remaining Nidaros sources for Grates honos, the Icelandic Stockholm fragment and Seqv 18 (Mi 106) both invert the order of the Fig. 33: Seqv 18/Mi 106 (Oslo, NRA Lat. fragm. 881-6, -7). Partial words from Grates honos hierarchia.

verses in strophe 7, which may indicate that this was a common feature in Nidaros for this sequence.

5.6.3. Liturgy of summer and early autumn Before taking a closer look at the other manuscript, the breviary-missal, it may be useful to take a look at how these manuscripts relate to the Nidaros ordinal, as shown in the table below. The table also includes two other thirteenth century manuscripts, the gradual Seqv 23 (Gr 26, Oslo, NRA Lat. fragm. 281) and the sequentiary Seqv 21 (Lat. fragm. 715).

207

Part II: Analysis Table 10: Selected manuscripts with sequences from June to August. Ordo Nidrosiensis (Gjerløw 1968, 358-387) In die ap. Petri et Pauli (29 June): Seq. Petre summe In comm. beati Pauli ap. (30 June): All. V. Magnus sanctus paulus vel Sancte paule Seq. Solennitas sancti Pauli ...175 In natalicio beati Olavi (29 July): All. V. Sancte Olave qui in celis vel Letabitur iustus Seq. Lux illuxit letabunda ...176 In die beati Laurentii (10 Aug): All. V. Levita laurentius Seq. Martyris eximii ...177 In die assumpt. beate Marie (15 Aug):178 Off. Gaudeamus omnes Gr. Propter veritatem All. V. Hodie maria virgo Seq. A rea virga Offert. Ave maria Com. Alma dei genitrix ... In octavis beati Laurentii (18 Aug): All. V. Beatus vir qui suffert Seq. Stola iocunditatis ... Sexta die assumptionis BMV: All. V. Succurre Seq. Postestate ... Vigilia beati Bartholomei ap. (24 Aug): All. V. Vox sancti Seq. Alle cantabile De sancto Audoeno (24 Aug)

Seqv 18173 (Mi 106)

Seqv 23 (Gr 26) lacuna

Seqv 38174 (Br-Mi 2) lacuna

Seqv 21 lacuna

x lacuna x x (V. Sancte Olave) x x x x lacuna x x x lacuna x

x x x in crast. ad matutinas179 lacuna

De sancto Hermete (27/28 Aug) 173

x -

For full survey of contents, see Part III: Catalogue. For full comparison with ON, see Ommundsen 2006b. 175 A lacuna in Seqv 18 goes from the last part of the feast of St. Paul (30 June) to the first part of that of St. Olav (29 July), through the octave of St. John (1 July) and the feasts of St. Swithun (2 July), St. Processus and Martinianus (2 July), St. Martin (4 July), Saints of Selja (8 July), Septem fratres (10 July), St. Benedict (11 July), In divisione apostolorum (15 July), St. Swithun’s translatio (15 July), St. Margareth (20 July), St. Mary Magdalen (22 July) St. Apollinaris (23 July), St. James (25 July) and Septem dormientes (27 July). 176 A lacuna in Seqv 18 goes from the last part of the feast of St. Olav to St. Lawrence’s introitus Confessio, through St. Abdon et Sennen (30 July), St. Peter ad vincula (1 Aug), St. Stephanus papa (2 Aug), inv. Sti. Stephani protomartyris (3 Aug), transfig. Domini (5 Aug), St. Donatus (7 Aug), St. Ciriacus (8 Aug). 177 Seqv 18 goes directly from the feast of St. Lawrence to the rubric In assumptione beate marie virginis (15 Aug). The ordinal prescribes the celebration of St. Tiburtius (11 Aug) and St. Ypolitus (13 Aug) in between. 178 The different elements are included for the Assumption of Mary, as they are relevant in the case of Seqv 23. 179 Seqv 38 has a proper lesson for St. Audoenus, not from the common as in the ordinal (Gjerløw 1968, 386). 174

208

5.6. Manuscripts secundum ordinem: Seqv 18 and 38 add, s. xiii med Festum beati augustini (27/28 Aug): Seq. Interni festi gaudia In exaltatione S. Crucis (14 Sept): Seq. Grates honos hierarchia

x lacuna x

The amount of space marked “lacuna” in the table clearly illustrates the frustrations of working with fragments, as the contents of the four selected manuscripts, which it would be interesting to compare, hardly overlap at all. In the table above I have only included the Alleluia-verses and the sequences, as they provide the most important information.180 The breviary-missal Seqv 38 add has three sequences for feasts coming in between those in the missal Seqv 18, namely: Stola iocunditatis (Octave of the feast of St. Lawrence, 17 Aug), Potestate non natura (Sixth day in the Octave of the Assumption of BMV, 20 Aug), and Alle cantabile (St. Bartholomew, 24 Aug). The texts and liturgical songs follow the Nidaros ordinal faithfully, although the readings for Audoenus are from his own vita and not from the Common of a confessor bishop as in the Nidaros ordinal. This is not a significant difference, as proper readings would seem more ideal than common and would be natural to include if available. The contents of Seqv 38 can therefore be said to follow the Nidaros ordinal rather faithfully. (For a more detailed discussion of the contents, see Ommundsen 2006b). It is worth noting that Seqv 38 add was “scored for reading”, i.e. that stress-marks were put in to facilitate the correct reading of the Latin text, a feature in medieval books drawn attention to by Leonard Boyle (1999). Although this feature is not unique, even in this small group of sources,181 it testifies to the practical use of the book.182 The liturgical books would include the temporale (going from the first Sunday of Advent, through major feasts like Christmas, Epiphany, Easter, the Ascension, 180

The Alleluiaverses from Trinity to Advent are relevant for the determination of rite, cf. Gjerløw 1968, 98. 181 Seqv 31 add (DRA, fr 3630), for instance, has a stress-mark over the word altérutrum (see Part III: Catalogue). 182 Words with stress-marks in Seqv 38 add are for instance invénit, benéfici, ádvene, valéret, úndique.

209

Part II: Analysis Pentecost, and Trinity to the 25th Sunday after Trinity) and the sanctorale, c. 150 saints feasts, about half of them with sequences in the Nidaros ordinal. Each major feast day in Seqv 38 add would have taken up at least four pages, or two leaves. The remaining parts of the breviary-missal probably come from a Sanctorale of approximately 225 leaves, or 28 quires (Ommundsen 2006b).183

Fig. 34: Seqv 38 add (Br-mi 3) (Oslo, NRA Lat. fragm. 668). At the top is the last line of the sequence Alle cantabile (the beginning of which is on the fragment labelled Sandaakers kat. LR pk. 34, C, Ryfylke 1651). Notice the small cap H in the fourth line from below. The invitatory Regem confessorum from the common of a confessor bishop, as prescribed in the Nidaros ordinal for St. Audoenus, is at the bottom of the page.

The missal, Seqv 18, probably had about 1 leaf per feast, and has been considerably smaller, with c. 150 leaves, or 19 quires, for the Sanctorale. The two books were not necessarily divided into one volume for the Temporale and one for the Sanctorale, 183

The calculation is based on the 150 feasts, the 75 with sequences going over two leaves, the 75 without sequences covering one leaf: 75 x 2 (leaves) + 75 x 1 (leaf) = 225 leaves.

210

5.6. Manuscripts secundum ordinem: Seqv 18 and 38 add, s. xiii med but could have had a winter volume, with the first half of the Temporale and winter portion of the Sanctorale, and a summer volume – a division which would certainly be more practical for the user.184

5.6.4. The authority of the Nidaros ordinal The full missal and breviary-missal in this chapter stand in contrast to the sequentiaries in the last chapter, in that they seem to strive towards following the ordinal. The minor differences, like the proper reading for the common, and the exclusion of some minor feasts, can be ascribed to a few generations of minor adjustments to the ordinal for practical use. One may imagine that with such a complex sequence repertory, it would be tempting for scribes of missals or breviarymissals to simply put in the sequence incipits, and expect the priest or cantor to supply the sequences from elsewhere. In this case, however, we have two very handy and portable books, perhaps used by priests who needed to travel between parishes, and the fact that the sequences were written out in full added to the functionality of the books. Without the sequences, an important function of the manuscripts, namely their completeness, would be lost. One question is if full sequentiaries in compliance with the ordinal were issued from Nidaros and spread along with the Nidaros ordinal. The clerics in the archbishopric must have been aware that in some remoter places it could be a challenge to have more than a hundred sequences of several different traditions available. None of the sequentiaries in this study gives any foundation to argue for the existence of a “Sequentiarium Nidrosiensis ecclesiae”. It is not impossible that some of the manuscripts in this study may have been copied from such sequentiaries, like Seqv 21 in Table 10 above for instance, but with only a couple of sequences one cannot assert this with any confidence. So where did the scribes of Seqv 18 and Seqv 38 add find the sequences which they entered into their respective full Mass books? Seqv 18 and Seqv 38 add have no overlapping sequences, which makes it impossible to see if the scribes of the two books had copied their sequences from a similar 184

I am grateful to Lori Kruckenberg for reminding me of this possibility.

211

Part II: Analysis “authoritative” exemplar, or if they were result of independent detective work done in Trondheim and Bergen (or Stavanger) respectively.

212

5.7. The St. Olav scribe: Seqv 45b add and Seqv 64 add, s. xiii²

5.7. The St. Olav scribe: Seqv 45b add and Seqv 64 add, s. xiii² The scribe called “the St. Olav scribe” is one of the few Norwegian scribes whose hand has been identified in several books. He wrote at least six books, two in Old Norse and four Latin liturgical books, at the end of the thirteenth century. His name is due to Lilli Gjerløw, coined from one of his books, “the legendary St. Olav’s saga” in Old Norse (Uppsala UB, De la Gardie 8 II) (cf. Gjerløw 1968, 35). Two manuscripts from this scribe’s hand are listed as sequence sources. The first is a missal/manual (Man 1) containing the sequence Eya recolamus (AH 53, no. 16) written out in full (Oslo, NRA Lat. fragm. 1028), while the second is an antiphoner (Ant 17) with a rubric containing the incipit of the sequence for St. Olav, Lux illuxit (Oslo, NRA Lat. fragm. 1030).

5.7.1. Six books from one hand That so many books remain from the hand of one scribe makes it possible to study the character of his work over time, as well as the character of the lay-out and decoration in his books. The St. Olav’s scribe is the second example in this study of a scribe writing in both Old Norse and Latin, therefore providing an opportunity to see the same scribe writing in different languages and genres – and what effect language may have had on his script. What is known about this scribe and his six known remaining works, their contents and connection to northern Norway is described by Lilli Gjerløw in her edition of the Nidaros ordinal (Gjerløw 1968, 35-38). In addition to St. Olav’s saga, the St. Olav scribe wrote a copy of King Magnus Lagabøte’s Lawbook, from which three fragments have survived.185 The Lawbook of Magnus (VI) Lagabøte (ruler of Norway 1263-1280) was introduced and accepted by the different regional law-districts 127476 as the valid law for all of Norway. The version of this Lawbook in the hand of the St. Olav scribe was probably adapted for the Frostathing (Gjerløw 1968, 36). Frostathing covered eleven counties surrounding Trondheim, and was originally

185

The fragments are found in the edition of NGL 2, 33-34, 88-90, 124-26 (cf. Gjerløw 1968).

213

Part II: Analysis seated at the peninsula Frosta, but later moved to Trondheim. The book was owned by the vicar Hans Olafssøn (d. 1621) in Bodø in northern Norway. It was later bought by the Danish collector Stephanius (d. 1652), and after his death acquired by the Swedish collector Magnus Gabriel De la Gardie, through whose collection it came to the Uppsala University Library. Among the liturgical works in Latin are a missal (Mi 27), a manual (Man 1), an antiphoner (Ant 17) and a breviary (Br 1). The antiphoner and breviary are also decribed in Gjerløw’s edition of the Nidaros antiphoner (Gjerløw 1979, 133, 250-1). The known works of the St. Olav’s scribe are as follows: Title or siglum

Signature

Extent

Language

The legendary saga of St. Olav

Uppsala, MS De la Gardie 8 II186

codex

Old Norse

King Magnus lagabøter’s Lawbook

Oslo, NRA Old Norse fragm. 47c

3 fragm.

Old Norse

Mi 27

Oslo, NRA Lat. fragm. 72

1 leaf

Latin

Man 1

Oslo, NRA Lat. fragm. 1028 a.o.

5 leaves

Latin

Ant 17

Oslo NRA Lat. fragm. 1030 a.o.

7 leaves

Latin

Br 1

Oslo NRA Lat. fragm. 1031

2 leaves

Latin

The leaf from the missal was used to bind an account from Senja in 1637 and is marked Mattis Holste.

187

The leaves from the manual were used to bind accounts

from Nordlands len (1628) and Senja (1629, 1631 and 1633) also marked Mattis Holste. The antiphoner and breviary was also used to bind accounts from Nordlands len, and it seems fairly certain that their secondary provenance was one or more churches in northern Norway. At least the missal and manual appear to having been available for the same bailiff at around the same time, and they were perhaps collected together. The proximity of the genres does make one ask if they could not have been from the same book. Both are in a quarto format with one column. The prickings in the upper margin for the vertical double lines framing the writing space appear very similar in the two 186 187

See the facsimile edition, Legendarisk Olavssaga (Holtsmark 1956b). Information from the unpublished database at the NRA.

214

5.7. The St. Olav scribe: Seqv 45b add and Seqv 64 add, s. xiii² manuscripts. In the missal 28 lines are visible, but, as the lower part of the leaf is cropped, it is not possible to know how many lines there were originally. The manual has 28 lines (disregarding the staves). Still, the writing space is of different width (13,5 cm for the missal, and 10,5 cm for the manual). It is therefore most likely that the missal and the manual are in fact two separate books. As all the six books by the scribe are in some way connected to the area from Trondheim to Bodø, this is most likely the area where he worked. The manual, according to Lilli Gjerløw, was penned for a church dedicated to St. Lawrence and another saint (Gjerløw 1968, 37), but this information is currently of limited use as most church dedications are now unknown. He either worked in Nidaros or further north, but his background was Nidaros, as Anne Holtsmark describes the language of the Olav’s saga as that of the Trondheim area (1956b, 7). This scribe did not necessarily work in a monastic setting or a larger scriptorium. None of his work seems to involve other hands or scribes, including rubrics or initials, as the initials in his books seem to have been made by the scribe himself. In other words, his requirements were merely a steady supply of parchment and ink to sustain his remarkable efficiency.

5.7.2. Not beautiful, but confident As Lilli Gjerløw pointed out, the St. Olav scribe did not write a beautiful hand. Still, this is not the awkwardness of the twelfth century, where scribes would try to copy or imitate the work of others, giving their products an air of “almost, but not quite”. The St. Olav’s scribe does not seem to have put much effort into the aesthetics of his writing, but given priority to productivity, readability and correctness. There is not a wide variation in his initials, which have a stylised form of the “sprouts” seen in ch. 5.3., either in red, blue or green with pen-decorations in black and yellow, or with blue pen-flourishing on red letters and vice versa, which looks more “up to date” (as in St. Olav’s saga).

215

Part II: Analysis

Fig. 35: Oslo, NRA Lat. fragm. 1031 (Br 1). This surviving leaf from a breviary is characteristic of the style of the St. Olav scribe. The initials are either clear red, an olive green, or pigeon blue. The written space measures 22 x 14/15 cm.

216

5.7. The St. Olav scribe: Seqv 45b add and Seqv 64 add, s. xiii²

Fig. 36: Above: Oslo, NRA Old Norse fragm. 47c, King Magnus Lagabøte’s Lawbook. The hand of St. Olav’s scribe did not change much when he wrote in Old Norse compared to when he wrote in Latin.

Fig. 37: Left: Oslo, NRA Lat. fragm. 1031 (Br 1). Compare the initial H in the lawbook and the breviary. The swiftness of the strokes in the initial may suggest that the St. Olav scribe was responsible for his own initials, as well as the rubrics. In the breviary he had access to a better quality red than when he was writing the lawbook.

217

Part II: Analysis In the case of the breviary, some of the plain monochrome initials have their models in the twelfth century (see for instance the decorative disc on the green initial N). As shown in the illustrations above, other initials on the breviary fragment are remarkably similar to those of the Old Norse law-book. The initals of the manual are also of the same type, but with slightly different colours, as the green is more malachite than olive. While an example of a Norwegian scribe writing below top line s. xiii med. has been noticed above, the St. Olav’s scribe, at least in the case of the breviary wrote above top line, even though he worked in the last half of the thirteenth century. It could be that this was not taken up in all parts of Norway, or that this particular scribe kept to his training, which may have been much earlier than when the breviary was written. The manuscripts of the St. Olav scribe are interesting as witnesses of the liturgy one stage removed from the Nidaros ordinal; to its contents have been added celebrations during the Octave of St. Olav and the feast of St. Francis (Gjerløw 1968, 38). The St. Olav scribe deserves his own study. What is interesting here is that language or genre does not seem to have any effect on the “Schriftbild” of this scribe, neither in regard to the execution of the script and its level of formality, nor in the decoration. The six books in all likelihood cover a fairly wide timespan. Still, the scribe seems to have held on to the same style throughout his career.

218

Conclusion for the thirteenth century

Conclusion for the thirteenth century While the twelfth century books had their origin in several different parts of Europe, the thirteenth century books in this corpus were mainly made in Norway. This is simply due to coincidence, as there is evidence of many imported thirteenth century manuscripts in the NRA. Although it in most cases is difficult to pinpoint an origin, there seems to be evidence of books from both eastern and western Norway as well as the Trondheim area. The earliest material from the thirteenth century appears to be the product of a scribe working in the north-west of Norway, heavily influenced from England in both script and initials. This scribe may have been one of the regular canons in St. John’s in Bergen. Another scribe, whose works are connected to each other for the first time in this thesis, seems to have had models from French script, as well as English. One hypothesis is that this scribe was one of the nuns at Gimsø, a monastery with a location close to the assumed secondary provenance of the manuscripts. A missal of a quality which is rarely seen in Norway is a valuable addition to our knowledge of the books which may have been produced here. This is heavily influenced from England in both script and decoration. Regarding book types, three sequentiaries from the first part of the thirteenth century do not seem to follow the Nidaros ordinal, and represent a classical type of AngloFrench sequentiary. These happen to have a provenance in the eastern parts of Norway, and of these at least two (Seqv 1 and 15) appear to be written after the Nidaros ordinal was issued. In these cases I have raised the question of the authority of the exemplar, since older sequentiaries continued to be copied in spite of the existence of an ordinal. The exemplar could have a stronger position in the case of sequentiaries than missals or breviary-missals, since sequences with their flexible repertories traditionally had to be collected from different sources. The parchment roll found in Lom in the 1970’s with a sequence is probably an example of an older tradition with its roots in the first decades of the sequences, when they would be copied on to rolls for rehearsals. In later times such parchment rolls would be an

219

Part II: Analysis efficient way of spreading single sequences without having to copy a whole book. If a sequence was missing from a particular book, it would have been easy to acquire it on a parchment role. Whether there was ever a thirteenth century “Sequentiarium Nidrosiensis ecclesiae”, or whether anyone ever saw the need for such a book, is unknown. The missal and breviary-missal copied in the mid thirteenth century in compliance with the Nidaros ordinal could have been evidence for one, as the scribes appear to have entered, in order, all the 111 sequences, or rather between 117 and 123 sequences (not counting the later marginal additions) in the ordinal as reconstructed by Kruckenberg (2006, 16). One may imagine that they collected the sequences from different sources, and not from one sequentiary, but as the contents of the surviving fragments do not contain the same sequences it is difficult to make any firm conclusion. However, the missal and breviary-missal do have much in common in their basic lay-out and format, and their level of formality and degree of decoration. The breviary-missal (Seqv 38 add/Br-Mi 3) using the colours red and green, was presumably produced on the west coast (Bergen?) and used in Stavanger. The missal (Seqv 18/Mi ) with the colours red and blue, with some letters in green, was probably written and used in the Trondheim area. Although the thirteenth century scribes wrote distinctive hands, most of them basically wrote an English type of script. At this time the tradition of writing books had been stable in most Norwegian towns for almost a century, and it does seem that a basic type of script had been established during this time. This means, as was the case for the English vernacular letterforms in the eleventh or twelfth century, an “English type script” of the first half or mid thirteenth century was a stable type, and a testimony to an earlier stage of English influence in the last half of the twelfth century. In the last half of the thirteenth century the St. Olav scribe is an interesting example of an efficient scribe, who wrote quickly and was prolific, although his work is not beautiful. Still, this is not the awkwardness of the twelfth century. The St. Olav’s

220

Conclusion for the thirteenth century scribe found his style and stuck to it, writing his non-beautiful script with energy and confidence whether he was writing in Old Norse or Latin.

221

Conclusion At the beginning of this study I set out to discover how fragments with sequences could shed light on a growing book culture and its European influences in the northern periphery. In the three parts of this thesis I have presented a selected corpus in the form of a catalogue, demonstrated one approach to the study of fragments from liturgical manuscripts also applicable for other fragment studies, and provided an analysis of selected fragments from twelfth and thirteenth century books. The results show that the Norwegian church related to a broad spectrum of competing influences throughout the twelfth century, which not only manifested itself in the repertory of sequences, but also in imported books and local book production. It is possible to see a pattern of development and change in the Norwegian scribal culture from the twelfth to the thirteenth century, shifting from a broad variety towards a larger degree of uniformity and confidence.

A book-culture of diversity Of the seventy Norwegian manuscripts with sequences presented in the catalogue, 11 are from the twelfth century, 26 are from the thirteenth, 8 from the fourteenth, 23 from the fifteenth and 3 from the sixteenth. In other words, 37 manuscripts, more than half of the corpus of seventy manuscripts, pre-date 1300. These 37 manuscripts from the twelfth and thirteenth century have been the primary focus of this study. 30 25 20 15 10 5 0 12th c.

13th c.

14th c.

15th c.

16th c.

222

Diagram 1: The chronological distribution of manuscript with sequences.

Conclusion The fragments from the eleven manuscripts from the twelfth century testify to a remarkable diversity, both in the early attempts at local book production and the origin of imported books. The imported books cover most of western Europe. The two oldest manuscripts in the study, from the first half of the twelfth century, testify to contact with France and with the Danish arch see of Lund. Imported books from the latter half of the century come from England and the Low Countries. Of the three local products from this period, two sequentiaries and one manual, all bear witness to different paths of influence: a small sequentiary has a script clearly influenced from England (and the binding of the book also seems to be done in an English manner), one sequentiary of a large format has a French or Anglo-french type of writing, and the manual is written in a German script (with a punctuation pointing to southern Germany). The manuscripts treated here support the indications of the earliest missals, namely that the influence from Europe was not a case of “first one, then the other”. Impulses from Germany through Hamburg-Bremen and also from Hirsau in southern Germany, probably through Lund, were not simply released or replaced by those from England. Several regions played a simultaneous role, although the balance seems to have shifted in favour of England during the course of the twelfth century. France has also clearly played an important role in the selected sequence repertory, local sequence composition, liturgy and script. The twelfth century evidence of French influence is most interesting, since France traditionally has not been considered in connection with Norwegian palaeography. From what time and how long this influence lasted is not certain, but France seems to have had a strong position at least from the first half of the twelfth century to the first half of the thirteenth. The twelfth-century corpus of manuscripts with sequences indicates that the “power struggle” between England, France and Germany extended beyond competing regional sequence repertories into the realm of script and books.

Among the 26 manuscripts in the thirteenth century, the majority seem to be local products. This does not mean that no manuscripts were imported in the thirteenth

223

century, they just are not part of this study. There might also be some manuscripts in the corpus imported from the other Scandinavian countries, but at present these are difficult to recognise. In the locally produced books of the thirteenth century there seems to have been a change towards a more established type of script. There is less variety, and “English-like” type hands seem to be the most dominant from the first decades onwards. The occurrences of peculiar punctuation and different types of a’s and g’s found in the twelfth century seem to be rarer in the thirteenth. The exception may be the leaves from Bratsberg (cf. Seqv 51 add), which appear to have features of a French type of script, with an un-English ampersand. To what extent it will be possible to characterize a “Norwegian” type of writing is not yet certain. The different scribes believed to be local show wide individual variety, and they are easier to contrast than to group, even though several of the characteristics of the script are the same. Features from Old Norse vernacular sometimes “contaminate” the Latin script even in the thirteenth century, and a certain old fashioned execution seems common among the local scribes. Still, the scribes of the thirteenth century seem more confident, even when they do not write a particularly beautiful hand. Some Norwegian scribes stand out, like the scribe from Bergen, possibly a regular canon at St. John’s, and the St. Olav scribe, presumably from Trondheim. It does seem possible to trace the contours of some “regional styles” also in Norway. Several of the manuscripts with a secondary provenance in western Norway around the middle of the thirteenth century have features in common, such as rather round letters and initials in red and green rather than red and blue. Regarding the manuscripts from after 1300, there is a big contrast between the fourteenth and fifteenth century. Even when allowing for some manuscripts to be assigned to the fifteenth which were actually written in the fourteenth century, it seems that there is still a very small number of fourteenth century manuscripts compared to the fifteenth. Espen Karlsen has pointed out that there was a drastic decrease in the number of missals from c. 1300, which could indicate a decreasing demand for liturgical books at this time as most churches had already been supplied

224

Conclusion with missals (Karlsen 2005, 151). Karlsen also mentions the Black Death as a factor. In Norway the plague spread from Bergen to the rest of the country in 1349, wiping out approximately fifty percent of the population. Seip discusses a decline in the number of extant manuscripts from about 1370, but suggests the possibility that the introduction of paper, which is not as durable as parchment, may have caused large losses (Seip 1954, 112). While Karlsen also finds a small number of missals from the fifteenth century, this is not the case for the sequence manuscripts, for with 23 manuscripts this is back to the level of the thirteenth century. When this tendency is not the same among the missals and the manuscripts with sequences (which also include missals), there could be at least two reasons for this: one is that Lilli Gjerløw when registering missals (of which there are large numbers) had a fondness for the older missals and gave these priority. We will know the answer to this when all the missals of the NRA are registered. Secondly, mere coincidence may have preserved a larger number of fifteenth century manuscripts with sequences, making the number of fifteenth century sequence manuscripts high compared to the missals. On the other hand, the relative number of manuscripts with sequences is low compared to the Swedish. That in the Norwegian material, there are almost as many thirteenth century as fifteenth century fragments stand in contrast to the Swedish material, where more than half of the total number of manuscripts with sequences are from fifteenth century manuscripts: while more than 250 of the total of 448 Swedish sequence manuscripts are dated to the fifteenth century, 115 are dated to the fourteenth and “only” 61 are dated to the thirteenth (Björkvall 2006, 48). There seems to be no evidence in Sweden of a drop in manuscript production after 1300 or after the Black Death. Regarding the Norwegian fifteenth century manuscripts, they are almost exclusively influenced from Germany, not England or France. Although one should be careful not to generalize the results of such a small study, the indications in the manuscript material with sequences are that the European points of contact had drastically changed between the twelfth and the fifteenth century.

225

() * +

,

'

! % &

Diagram 2: The origin of twelfth century manuscripts with sequences

#

" $

! "

Diagram 3: The origin of fifteenth century manuscripts with sequences

The fifteenth century manuscripts also testify to a change in the attitude towards sequences which led to the dramatic change in the sequence repertory of the printed Missale Nidrosiense (1519). Whether these changes were the result of a gradual development or decisions made in connection with the printed edition of the missal is a problem for further study. The Icelandic material suggests that the Icelandic mass books were largely in agreement with the Nidaros ordinal in the number and selection of their sequences until at least the last quarter of the fifteenth century (Attinger 2006, 176). As regards the Norwegian material, this remains to be investigated.

226

Conclusion Sequences in context Sequences are found in a large number of liturgical book genres. The seventy manuscripts are distributed on the following genres: 37 sequentiaries, 10 manuals, 7 missals, 4 graduals, 4 antiphoners, 3 breviary-missals, one antiphoner-gradual, one roll and three of unknown book genre.

3

3 11 Sequentiary

4

Manual Missal

4

Gradual 37

7

Antiphoner Breviary-missal Unknown Antiphoner-gradual Roll

10

Diagram 4: The genre distribution of the manuscripts with sequences

Regarding the book types and formats the variety continues. In the twelfth century there are examples of sequences in books ranging from a small, handy breviarymissal to a huge folio-format missal. The sequentiaries range from the “classical” small sequentiaries to a large folio format sequentiary. While the small English sequentiary probably carried a repertory of Anglo-French sequences, the large sequentiary seems to be a local copy of an exemplar with a German repertory built on Notker-sequences. In other words, among the sequentiaries in the medieval churches some would have reflected an Anglo-French, some a German sequence tradition.

227

Some manuscripts would also have contained a merging of the Anglo-French and German traditions. By chance, there is a twelfth century local product displaying an example of how a scribe in the periphery could relate to two competing influences: he copied both into the same manuscript. An Anglo-French Christmas-sequence and a German Notker-sequence for Christmas which are not found earlier in the same country, let alone in the same manuscript, stand side by side in a Norwegian sequentiary. This is one example of how important it can be to look at sequences in context, as part of a manuscript, and not simply as single entities. The same sequentiary also testifies to the dominating role the English impulses had over the German in the second half of the twelfth century, not only through its English appearance, but also through putting the German sequence in second place, giving the Anglo-French sequence higher priority, a feature found later in the Nidaros ordinal. Among the fragments from sequentiaries is no sign of a “Sequentiarium Nidrosiensis Ecclesiae” in compliance with the Nidaros ordinal. There may very well have been such a type, which could have served as a source for the sequences in the thirteenth century missal and breviary-missal which presumably contained a full set of sequences. Medieval scribes related to several traditions and demands. On the one hand there was probably a large number of sequentiaries following one or the other tradition. On the other hand there was the demands of the Nidaros ordinal. As an explanation why not all the sequentiaries produced locally in the thirteenth century follow the Nidaros ordinal, I have suggested the presence of a factor which may be called “the authority of the exemplar”. No liturgical book would be copied without one or more exemplars, and a complete book would hold a certain authority. This might be the reason why scribes in some cases seem to have copied an exemplar with only a slight adaptation to the local rite (like the inclusion of St. Olav’s sequence), rather than rework and combine different sequentiaries to fit the Nidaros ordinal. The dynamic and tension between the authority of the exemplar and the authority of the ordinal would probably be more evident in a book genre like the sequentiary than a missal.

228

Conclusion There would probably be higher demands for a missal to follow the ordinal to the letter, to ensure the correct celebration of the mass (although not all thirteenth century missals are secundum ordinem either). A sequentiary was more flexible, and sequences could easily be supplemented later with rolls or single leaves (which later could be combined in one book, if desired). The presence of a parchment roll in the corpus is significant, and a reminder that the codex was not the only used medium for liturgical music. It seems that while the North was “Europeanized” the lack of tradition gave a certain freedom, and the evidence of both liturgy and the liturgical books indicates that this freedom was used to select that which was preferred at a given time. The evidence of both sequences and script clearly shows that no “pact of obedience” was formed with any one European centre, although certain smaller Norwegian centres, institutions or people may have had stronger links to some places than others. In an earlier chapter I referred to the North as a melting pot, and the particular “northern blend” seems to have been the result of a combination of initial reception and training, and, after a time, active and conscious selection and production.

Ad fragmenta! I have argued that the remaining manuscript material represents less than 1% of what existed in Norway at the time of the Reformation. Although this is unfortunate, the remains of c. 1200 manuscripts are in the fragment collection at the NRA, most of which liturgical, to a large degree unstudied. In this thesis I have dealt with seventy manuscripts, of which fragments from 58 are in the NRA, while 12 are in other collections. So although little remains compared to what once was, the fragments still represent such a large quantity of material that it will be years before scholars have an overview of the majority of the material. As I have shown in this thesis, fragments from liturgical manuscripts are in fact most suitable for the study of Norwegian book culture. Due to the difficulties connected to fragment studies of this type, I have described different approaches and

229

demonstrated how a type of genre-study can be done. As my purpose was not only to separate imported books from local, but also to detect layers of scribal influence in the local products, it was necessary to form a specific method, identifying a moderate number of single characteristics to look for. This method is not meant to be used mechanically, but with great care, as there are few absolutes when working with fragments from medieval manuscripts. Scribes and books were highly movable and some scenarios are untraceable. In addition, each evaluation of a manuscript is based on personal opinion, and it is not possible to exclude a certain error in judgement. These insecurities will always be a factor in fragment and manuscript studies, but it will be possible to achieve a larger degree of confidence when considering contents, script and the physical aspects of the book together. Regarding script, the goal is to develop an “eye” for identifying their regional features, but the first step on the road to this goal, whether we like it or not, is lists of significant features. We are well on the way to achieve a better understanding of the past, not least because borders between disciplines are lowered. To get a full understanding of the fragments from a book culture which chance has brought down to us, we need the expertise of musicologists, historians, art historians, Latin and Old Norse philologists and palaeographers – and also the valuable experience of archivists and librarians. Latin and Old Norse – books or book fragments and documents – are parts of the same system. It is of vital importance that we keep a broad horizon and learn from international collaboration. Hopefully we will in turn be able to contribute to the knowledge about the mechanisms at work when Christian Europe expanded and embraced the outer edges in the north and in the east. The manuscripts brought northwards from several different regions of Europe should be subject to further study, and so should the manuscripts which testify to local book production. This coin has been buried in the ground for too long. Although a lament of the losses is natural, it is now time to look forward and make use of this opportunity which fortune has left us. “Ad fontes” could in this case be

230

Conclusion rephrased as “ad fragmenta”. I hope that this study will be followed by others using the fragments in the NRA, and that each will contribute further to our knowledge about medieval book culture.

231

232

Indices

Indices

233

Manuscripts

ARENDAL Aust-Agder kulturhistoriske senter: AA 72: seglremmer (Seqv 54 add): see catalogue AA 4981-2: fragment (Seqv 55 add): see catalogue AA skinnbrev (ed. DN 7, 441): 120 BERGEN University Library MS 1550, 7 (Seqv 56a add): see catalogue State Archives Rosendal (formerly UBO, Dep. no. 269:7) (Seqv 17): see catalogue BERLIN Kupferstichkabinett Psalter of Margrete Skulesdotter: 72 CAMBRIDGE Corpus Christi College 422 (The Red book of Darley): 55 University Library Add. 710 (The Dublin Troper): 196 COPENHAGEN The Arnamagnæan Collection AM 98 8° II: 80 AM 241 A III Acc. 7b: 234 AM 241b fol. IV: 80 AM 243b fol. (The king’s mirror): 71, 72, 86 AM 619 4° (The Old Norwegian Homily Book): 71, 72, 86, 157, 160, 171 The Royal Library Add 120 (Seqv 59a add): see catalogue GKS 1005 fol. (Flateyjarbók): 118 GKS 1347 4° (The Munkeliv Gospel Book): 70, 72, 195 GKS 1606 4° (The Kristina Psalter): 72 NKS 32 8° (Seqv 59b add): 70, 72, 147 ff. and catalogue NKS 133 f, 4° (Seqv 60 add): 70, 72, 147 ff. and catalogue Thott 110 8°: 70, 72, 147

234

Indices

The National Archives Fragm. 3031-32: Bahus a 1622-23 (Seqv 61a add): see catalogue Fragm. 3071-72: Bahus c 1622 (Seqv 61b add): see catalogue Fragm. 3630-31 LR Kristianopel 16 (Seqv 31 add): 122 and catalogue LONDON British Library Add. 11669: 207 Add. 74236: 157 Arundel 83: 199 Harley 2961: 55 LUND University Library Medeltidshandskrift 5: 125 Medeltidshandskrift 6: 127 MUNICH Bayerische Staatsbibliothek clm 13845: 207 OSLO

Deichman Library Aslak Bolt’s Bible: 72 Museum of Cultural History, University of Oslo C 34738/52 (Seqv 56 add, the Lom parchment roll): 198 and catalogue The National Archives (old numbers) Lat. fragm. 28: 127 Lat. fragm. 72 (Mi 27): 214 Lat. fragm. 97 and 98: 185 Lat. fragm. 103: 185 Lat. fragm. 107: 185 Lat. fragm. 108: 185 Lat. fragm. 235 (Seqv 31 add): 122 ff. and catalogue Lat. fragm. 236 (Seqv 30): 142 ff. and cataogue Lat. fragm. 251 (Seqv 29/Br-mi 2): 114 ff. and catalogue Lat. fragm. 256 (Mi 133): 134 ff. Lat. fragm. 261 (Seqv 32 add): see catalogue Lat. fragm. 281 (Seqv 23/Gr 26): 207 and catalogue Lat. fragm. 284 (Seqv 3): see catalogue Lat. fragm. 294 (Seqv 33 add): 181 ff. and catalogue Lat. fragm. 296: 128 Lat. fragm. 354 (Mi 133): 134 ff. Lat. fragm. 366 (Seqv 2): 193 ff. and catalogue 235

Lat. fragm. 418 (Seqv 1): 17, 52, 193 ff. and catalogue Lat. fragm. 419 (Seqv 15): 193 ff. and catalogue Lat. fragm. 427 (Seqv 10): see catalogue Lat. fragm. 433 (Gr 27): 177 ff. Lat. fragm. 452 (Lec-Mi 4): 173 ff. Lat. fragm. 471 (Seqv 16): 140 and catalogue Lat. fragm. 489 (Seqv 33b add): 181 ff. and catalogue Lat. fragm. 497 (Seqv 11): 134 ff. and catalogue Lat. fragm. 506 (Seqv 8): see catalogue Lat. fragm. 507 (Seqv 8): see catalogue Lat. fragm. 511 (Seqv 33b add): see catalogue Lat. fragm. 532 (Seqv 25): 188ff. and catalogue Lat. fragm. 533 (Seqv 9): see catalogue Lat. fragm. 553 (Seqv 34 add): see catalogue Lat. fragm. 571 (Seqv 35 add): see catalogue Lat. fragm. 573 (Seqv 36 add): see catalogue Lat. fragm. 627 (Seqv 5): 52 and catalogue Lat. fragm. 634 (Seqv 19): see catalogue Lat. fragm. 664 (Seqv 37 add): see catalogue Lat. fragm. 668 (Seqv 38 add): 201ff. and catalogue Lat. fragm. 694 (Seqv 28): see catalogue Lat. fragm. 715 (Seqv 21): 207 and catalogue Lat. fragm. 750 (Seqv 39a add): see catalogue Lat. fragm. 764 (Seqv 39b add): 157 ff. and catalogue Lat. fragm. 776 (Seqv 40 add): 129 ff. and catalogue Lat. fragm. 797 (Seqv 41 add): see catalogue Lat. fragm. 798 (Seqv 42 add): see catalogue Lat. fragm. 803 (Seqv 43 add): see catalogue Lat. fragm. 813 (Seqv 44 add): see catalogue Lat. fragm. 843 (Seqv 24): see catalogue Lat. fragm. 870 (Seqv 27): see catalogue Lat. fragm. 881 (Seqv 18/Mi 106): 201 ff. and catalogue Lat. fragm. 882 (Seqv 18/Mi 106): 201 ff. and catalogue Lat. fragm. 911 (Seqv 6): see catalogue Lat. fragm. 932 (Seqv 18/Mi 106): 201 ff. and catalogue Lat. fragm.952 (Seqv 12): see catalogue Lat. fragm. 978 (Seqv 14): see catalogue Lat. fragm. 980 (Seqv 4): see catalogue Lat. fragm. 984ab (Seqv 45 a add): see catalogue Lat. fragm. 986 (Seqv 13): see catalogue Lat. fragm. 1007 (Seqv 22a): see catalogue Lat. fragm. 1018 (Ant 7): 158 ff. Lat. fragm. 1027 (Ant 7): 158 ff. Lat. fragm. 1028 (Man 1): 213 Lat. fragm. 1029 (Seqv 46 add): 17 and catalogue Lat. fragm. 1030 (Seqv 64 add/Ant 17): 213 and catalogue Lat. fragm. 1031 (Br 1): 214 Lat. fragm. 1034 (Ant 7): 158 ff. 236

Indices Lat. fragm. 1075 (Seqv 47 add): see catalogue Lat. fragm. 1088 (Seqv 7): see catalogue Lat. fragm. 1101 (Seqv 4): see catalogue Lat. fragm. 1114 (Seqv 20): see catalogue Lat. fragm. 1120 (Seqv 20): see catalogue Lat. fragm. 1046 (Seqv 33b add): see catalogue Lat. fragm. box 44, C (Seqv 48 add): see catalogue Lat. fragm. pk. 52 (Seqv 26): see catalogue Lat. fragm. Sandaakers kat. pk. 142 (B.hus)(Seqv 53 add): 129 ff. and catalogue Lat. fragm. Sandaakers kat. pk. 157 (Seqv 51 add): 172 ff. and catalogue Lat. fragm. Sandaakers kat. pk. 692 (Seqv 50 add): see catalogue Lat. fragm. Sandaakers kat. pk. 717 (Seqv 49 add): see catalogue Lat. fragm. Sandaakers kat. pk. 718 (Seqv 52 add=Seqv 20): see catalogue Lat. fragm. XIX (Seqv 22b): see catalogue Old Norse fragm. 47c: 214 The National Library Ms.8° 102 (The Kvikne Psalter): 70, 72 OXFORD Bodleian Library Bodley 775: 126 University College MS 148: 196 PARIS

Bibliothèque Nationale Lat. 1107: 157, 196

PRAGUE Prague Castle, Archiv Prazskeho hradu Psalter from Munkeliv, Bergen: 72 REYKJAVIK The National Library fragm. 19: 80 The National Museum no. 3411: 80 SANKT GALLEN Stiftsbibliothek MS 376: 143 SCHAFFHAUSEN Ministerialbibliothek Ms. 95: 143 237

STAVANGER Stavanger museum Bookbinding on Chr. Scriver 1692 (Seqv 62 add): see catalogue State Archives Fragm. 1-3 (Mi 35): 184 Fragm. 13 (Seqv 38 add/Br-Mi 3): 201 and catalogue STOCKHOLM The Royal Library Pap. 18 4°: 207 Perg. 15 4° (The Icelandic Homily Book): 161 The National Archives Fr 7134: 186 TRONDHEIM Gunnerus Library (the University Library) Fragm. 9 (Seqv 31 add/Mi 80): 122 and catalogue Fragm. 29: see catalogue UPPSALA University Library De la Gardie 8 II: 72, 86, 213

238

Indices

Gjerløw’s alternative sigla for manuscripts with sequences

Ant 17 (= Seqv 64 add) Ant 18 (= Seqv 45a add) Ant 34 (= Seqv 43 add) Br-Mi 2 (= Seqv 29) Br-Mi 3 (= Seqv 38 add) Br-Mi 5 (= Seqv 33a add) Gr 10 (= Seqv 17) Gr 20 (= Seqv 41 add) Gr 24c (= Seqv 24) Gr 26 (= Seqv 23) Gr 44 (= Seqv 49 add) Gr 45 (= Seqv 50 add) Gr 55 (= Seqv 4) Man 1 (= Seqv 45b add) Man 19 (=Seqv 22a) ManNor B (= Seqv 60 add) ManNor F (= Seqv 59b add) Mi 38 (= Seqv 39b) Mi 49 (= Seqv 46 add) Mi 69 (= Seqv 44 add) Mi 80 (= Seqv 31 add) Mi 106 (= Seqv 18) Mi 107 (= Seqv 33b) Mi 112 (= Seqv 22b) Mi 133 (= Seqv 11?)

239

Sequences in Norwegian and Icelandic fragments and manuscripts Title

AH

Egg. Gj. ON

MN Manuscripts

Ab arce siderea Ad celebres rex celice

8:5 7:178 53:190 54:88 53:241

x x

x

Agni paschalis esu Agone triumphali Alle cantabile sonet

53:50 53:229 8:137 53:129

x x x

x x

x x

Alle celeste nec non

7:98 53:97

x

x

x

Alleluia celica resonant Alleluia nunc decantet Alludat letus ordo

x x x

Auro virginum intestus Ave Maria gratia plena

53:19 54:86 7:176 53:189 53:87 8:96 10:211 7:79 53:76 55:326 7:107 53:106a 54:216

Ave maris stella Dei Ave mundi gloria Ave mundi spes Maria

Ad laudes salvatoris Adest nobis dies alma

Alma chorus domini Alma dei genitrix Alme concrepent Almiphona iam gaudia Aquas plenas Aurea (A rea) virga

x

x

x

x x x x x

x x

x

x

x

x

x x

x

8:92 54:254 54:217

x x x

x

Ave preclara maris stella Benedicta es celorum r.

50:241 54:252

x x

x x

x

x

Benedicta semper sancta

7:95 52:81

x

x

x

x

Benedicta sit beata tr.

7:96 53:81b

x

x

x

188

Iceland only Seqv 1: Lat. fragm. 418 Seqv 5: Lat. fragm. 627 Seqv 61a add: DRA, Bahus a 1622-23 Seqv 1: Lat. fragm. 418 Seqv 3: Lat. fragm. 284 Seqv 30: Lat. fragm. 236 Seqv 61b add: DRA, Bahus c 1622? Seqv 21: Lat. fragm. 715 Seqv 38 add (Br-Mi 3): Lat. fragm. Sk 34 and 668. Seqv 1: Lat. fragm. 418 Seqv 15: Lat. fragm. 419 Seqv 33 add (Br-Mi 5): Lat. fragm. 294 Iceland only, see Eggen 1968 I, p. L MN only Iceland only Seqv 29: Lat. fragm. 251 Iceland only Iceland only Iceland only

x x

188

Seqv 57 add: MCH, UiO C 34738/52 Seqv 1: Lat. fragm. 418 Seqv 23 (Gr 24): Lat. fragm. 281 Iceland only, see Eggen 1968 I, p. L Seqv 10: Lat. fragm. 427 Seqv 43 add (Ant 34): Lat. fragm. 803 Iceland only Iceland only Seqv 1: Lat. fragm. 419 Seqv 19: Lat. fragm. 634 Seqv 43 add (Ant 34): Lat. fragm. 803 Seqv 60 add (Man Nor B): Cph KB, MS NKS 133f 4° Seqv 61a add: DRA, Bahus a 1622-23 Seqv 60 add (Man Nor B): Cph KB, MS NKS 133f 4° Seqv 5: Lat. fragm. 627 Seqv 34 add: Lat. fragm. 553 Seqv 58 add: Trondh. GB, fragm. nr. 29 Seqv 1: Lat. fragm. 418 Seqv 12: Lat. fragm. 952 Seqv 13: Lat. fragm. 986

The list is a combination of the sequences found in Eggen (1968) and the sequences discovered after Eggen’s death in 1957 by Gjerløw and presented in an unpublished list titled “Tilveksten av sekvenser funnet etter 1957” (“Additional sequences found after 1957”), available at the NRA. Sequences found only in Icelandic fragments are in cursive script. For fragments in the NRA in Oslo, the place is not indicated. When a fragment is found elsewhere, the collection is specified.

240

Indices Title

AH

Egg. Gj. ON

MN Manuscripts

Celeste organum hodie

7:35 54:1 50:267

x

x

Celi enarrant gloriam Celica resonant Celsa lux Syon Celsa pueri concrepent Celum deus inclinavit Christe tui famuli Christi baptista Christi hodierna Christo inclita candida Clare sanctorum senatus

Claris vocibus inclita Comitis generosi Concentu parili Concepcio Marie virginis Congaudent angelorum Congaudentes exultemus

Copiose caritatis De profundis tenebrarum Deus in tua virtute Dilecto regi virtutum Dixit dominus ex Basan Ecce pulchra Eia Christi concio Eia musa Eia recolamus

Epiphaniam domino Exulta celum Exultet in gloria Festa Christi omnis

7:21 53:19 54:69 53:162 54:262 9: 242 7:23 53:17 7:118 53:115 53:228

7:104 53:101 53:99 54:188 53:104 54:66

28:29b 53: 122 40:258 50:269 7:116 53:114 7:77 53:75 53:16

7:37 53:28 9:243 53:29

x

x x

x

x

x

x x x x

x

x x

x

x

x

x

x

x

x

x x x

x

x

x x x x

Seqv 28: Lat. fragm. 694 Seqv 39 add: Lat. fragm. 750 Seqv 53 add: Sand. kat LR pk 142

x

x

x x

x x

x

x x

x x

x

x

x x

x

x

x

x

x

x

x

x

x

x

x x x

x x x

241

Iceland and MN only

Iceland only Seqv 2: Lat. fragm. 336 Seqv 7: Lat. fragm. 1088 Seqv 8: Lat. fragm. 506, 507 MN only Iceland only Seqv 39a add: Lat. fragm. 750 Seqv 2: Lat. fragm. 336 Seqv 6: Lat. fragm. 911 Seqv 1: Lat. fragm. 418 Seqv 16: Lat. fragm. 471 Seqv 1: Lat. fragm. 418 Seqv 5: Lat. fragm. 627 Seqv 12: Lat. fragm. 952 Seqv 61b add: DRA, Bahus c 1622 Iceland only Iceland only MN only Iceland only Seqv 36 add: Lat. fragm. 573 Seqv 1: Lat. fragm. 418 Seqv 5: Lat. fragm. 627 Seqv 15: Lat. fragm. 419 Seqvn 16: Lat. fragm. 471 Seqv 42 add: Lat. fragm. 798 Seqv 47 add: Lat. fragm. 1075 Seqv54 add: Arendal, AA 72 Iceland only Iceland only Seqv 47 add: Lat. fragm. 1075 Seqv 28: Lat. fragm. 694 Seqv 54 add: Arendal, AA 72 Seqv 55 add: Arendal, AA 4981-2 Seqv 2: Lat. fragm. 336 Seqv 15: Lat. fragm. 419 Iceland only Iceland only Seqv 2: Lat. fragm. 336 Seqv 4: Lat. fragm. 1101 Seqv 35 add: Lat. fragm. 571 Seqv 45b add (Man 1): Lat. fragm. 1028 Seqv 44 add (Mi 69): Lat. fragm. 813 Seqv 8: Lat. fragm. 504, 507 Iceland only Seqv 5: Lat. fragm. 627 Seqv 55 add: Arendal, AA 4981-2 Seqv 59a add: Cph KB, MS Add 120 fol.

Title

AH

Egg. Gj. ON

Fulgens preclara

7:44 53:35

x

Fulget mundo celebris Gaude Maria virgo Gaude pia Magdalena Gaudete vos fideles Gaudia mundo (inc. only) Gloria laus et honor Grates nunc omnes Grates honos hierarchia Gratulemur dulci prosa Hanc concordi famulatu Hec est sancta sollemnitas Hodierne lux diei cel. Hodierne lux diei sacram. Imperatrix gloriosa Innocentem te servavit In sapientia disponens In superna ierarcha Interni festi gaudia

53:85

x x x x

9:920 54:3 37:1 50:117 53:10 50:239 42:215 53:215 53:56

x x

54:219 54:173

x x

54:221 54:116 55:74

x x x x x

Inviolata, intacta et c.189

x

x

x

Seqv 9: Lat. fragm. 533 Seqv 11: Lat. fragm. 497 Seqv 62 add: Stavanger mus. Scriver Iceland only Iceland only, see Eggen 1968 I, p. XLIX Iceland only Seqv 9: Lat. fragm. 533 Seqv 39b add (Mi 38): Lat. fragm. 764

x

x

x x x

x

x x x

x

x

x x

x

x

x

Iocundare plebs fidelis

55:7

x

x

Iohannes Iesu Christo

53:168

x

x

Iste confessor

x

Lauda Syon salvatorem

2:101 51:118 42:312 54:165 7:13 53:4 50:385

x

Laudes Christo canamus Laudes Christo redempti Laudes crucis attollamus

9:322 53:45 54:120

x x x

Laudes debitas deo Laudes deo devotas Laudes salvatori Laurenti David magni Laus devota mente Laus tibi Christe cui sapit

54:62 54:14 53:36 53:173 40:379 53:156

x x x x x x

Iubar mundo geminatur Iubilemus cordis voce Iubilemus omnes una

MN Manuscripts

x x

x

x x

x

x x x

x x

x x x x

189

x

x

Iceland only, see Eggen 1968 I, p. LIII MN only Seqv 18: Lat. fragm. (id. Ommundsen) Seqv 28: Lat. fragm. 694 (id. Gjerløw) Seqv 35 add: Lat. fragm. 571 Seqv 59b add (ManNor F): Cph KB, MS NKS 32 Iceland only MN only Iceland only Iceland only Seqv 9: Lat. fragm. 533 Iceland only Seqv: 21: Lat. fragm. 715 Seqv 61b add: DRA, Bahus c 1622 Seqv 20: Lat. fragm. 1120 Seqv 22a (Man 19): Lat. fragm. 1007 Seqv 64 add: Lat. fragm. 780 Seqv 28: Lat. fragm. 694 Seqv 51 add: Sand. kat LR pk 157 Seqv 1: Lat. fragm. 418 Seqv 4: Lat. fragm. 1101 Seqv 35 add: Lat. fragm. 571 Iceland only, see Eggen 1968 I, p. L Iceland only MN only Seqv 1: Lat. fragm. 418 Seqv 2: Lat. fragm. 336 Seqv 12: Lat. fragm. 952 Seqv 34 add: Lat. fragm. 553 Seqv 18 (Mi 106): Lat. fragm. 881 Seqv 30: Lat. fragm. 236 Seqv 1: Lat. fragm. 418 Seqv 5: Lat. fragm. 627 Seqv 13: Lat. fragm. 986 Seqv 20: Lat. fragm. 1120 Iceland and MN only Seqv 1: Lat. fragm. 418 Seqv 59a add: Cph KB, MS Add 120 fol. Seqv 1: Lat. fragm. 418 Seqv 12: Lat. fragm. 952 (id. Gjerløw) Iceland only

”The Roman Catholic Church does not reckon this song among sequences, but it has all the characteristics of one, and is therefore included here” (Eggen 1968 I, 165)

242

Indices Title

AH

Egg. Gj. ON

Laus tibi Christe qui es Leta mundus sit iocunda Letabundus exultet

50:268 54:2

x

Lumen vite sanctis dat. Lux fulget hodierna Lux illuxit letabunda...lux est nobis Lux illuxit letabunda...lux illustris

10:214 8:6 -

x x x

x x

42:302

x

x

Magnus deus in universa Mane prima sabbati

7:201 53:220 54:143

x

x

x

Martiris egregii Martiris eximii Mera promat uranica Mirabilis deus Missus Gabriel de celis

55:340 55:218 53:231 54:192

x x x x x

Mundi etate Mundo Christus oritur Nativitas Marie virginis Nato canunt omnia

8:113 40:349 54:188 7:31 53:24

x x x

Natus ante secula Nunc luce alma Nunc nobis succurre O alma trinitas O beata beatorum O mira domini pietas Omnes gentes plaudite Omnes sancti seraphim Omnipotentem semper Omnis fidelium ecclesia Organicis canamus

53:15 37:276 9:221 7:97 55:14 54:152 53:112 50:122 40:191 7:152 53:242 53:46 9:930

x x

Petre summe Christi pastor

53:210

x

Plangas cum lacrymis Post partum virgo Maria Postquam calix Babilonis cf. Predicasti dei care

11:79 53:109 55:272

x x x

Pangamus creatoris Pangat nostra concio

x

x x x x x x x x

x

x

x

x

MN Manuscripts

x

x

Seqv 1: Lat. fragm. 418 Seqv 13: Lat. fragm. 986 Seqv 18 (Mi 106): Lat. fragm. 881 Seqv 1: Lat. fragm. 418

x

Seqv 1: Lat. fragm. 418 Seqv 4: Lat. fragm. 980 Seqv 13: Lat. fragm. 986 Seqv 52 add: Lat. fragm. Sk 718 Seqv 61b add: DRA, Bahus c 1622 Iceland only Seqv 18 (Mi 106): Lat. fragm. 881 Iceland only Iceland only Seqv 10: Lat. fragm. 427 Seqv 19: Lat. fragm. 634 Seqv 61a add: DRA, Bahus a 1622-23 Seqv 28: Lat. fragm. 694 (id. Gjerløw) Seqv 6: Lat. fragm. 911 MN See Concepcio Seqv 1: Lat. fragm. 418 Seqv 2: Lat. fragm. 336 Seqv 53 add: Lat. fragm. Sk 142 Seqv 53 add: Lat. fragm. Sk 142 Seqv 40 add: Lat. fragm. 776 Seqv 28: Lat. fragm. 694 (id. Gjerløw) Iceland only MN only Seqv 44 add (Mi 69): Lat. fragm. 813 Iceland only Seqv 31 add (Mi 80): Lat. fragm. 235 Iceland only, see Eggen 1968 I, p. LIII Iceland only Seqv 3: Lat. fragm. 284 Seqv 24 (Gr 24c): Lat. fragm. 843 Iceland only Seqv 6: Lat. fragm. 911 Seqv 13: Lat. fragm. 986 Seqv 25: Lat. fragm. 532 Seqv 5: Lat. fragm. 627 Seqv 8: Lat. fragm. 507 Seqv 18 (Mi 106): Lat. fragm. 881 Seqv 36 add: Lat. fragm. 573 Seqv 39 add: Lat. fragm. 750 Iceland only, see Eggen 1968 I, p. LIII Iceland only MN only? See Eggen 1968 I, 222-7.

x x x

x

x

x

x

x

x x x

x x x

x

x

x

x x x x x x

x

x

x

243

Iceland only Seqv 5: Lat. fragm. 627 (id. Gjerløw) Seqv 4: Lat. fragm. 1101 Seqv 5: Lat. fragm. 627 Iceland only Iceland only Iceland only

Title

AH

Egg. Gj. ON

Potestate non natura Precursorem...en baptista Predicasti dei care, cf. Postquam calix babylonis Pro nobis ora (inc.) Profitentes unitatem Prome casta concio

54:96 42:252 55:272

x x

Psallat ecclesia mater dec. Psallat ecclesia mater ill. Pure mentis gaudia Quare fremuerunt gentes Quicunque vult salvus Qui regis sceptra Regem celi cantico Regis et pontificis Regnantem sempiterna Resonet sacrata iam Resurrexit leo fortis Rex omnipotens Rex regum deus noster Sacerdotem Christi Martin. Sacrosancta hodierna

37:321 54:161 7:47 53:49 37:306 53:227

x

x

Seqv 38 add (Br-Mi 3): Stav. SA fr. 13 Iceland only Seqv 13: Lat. fragm. 986

x x x

Seqv 33b add (Mi 107): Lat. fragm. 489 Seqv 58 add: Trondh. GB, fragm. nr. 29 Seqv 62 add: Stavanger mus., Scriver

x x

x

40:138 -

x

54:163

x

7:9 53:3 8:16 7:7 53:2 7:78 53:74 7:72 53:66

x

x

x x x

x

x

x x x

x

54:30

x

x

x

Salve crux arbor vite Salve crux sancta

7:4 53:1 54:121 53:82

Salve lignum sancte cruc. Salve martyr alme Salve mater salvatoris Sancta cunctis letitia Sanctarum virginum Sancte Syon assunt

54:122 42:260 54:245 40:24 55:33

x

x

Sancti baptiste

53:163

x

Sancti spiritus assit

53:70

x

x

x x

x

x

Seqv. 1: Lat. fragm. 418 Seqv 58 add: Trondh. GB, fragm. nr. 29 Seqv 1: Lat. fragm. 418 Seqv 2: Lat. fragm. 336 Iceland only Seqv 13: Lat. fragm. 986 Seqv 1: Lat. fragm. 418

x

Seqv 37 add: Lat. fragm. 664

x

x x

x x

Iceland only Seqv 17 (Gr 10): Bergen, SA, Rosendal Seqv 24: Lat. fragm. 843 Seqv 54 add: Arendal, AA 72 Iceland only Seqv 25: Lat. fragm. 532 (id. Gjerløw)

x

x

53:243 53:181

Salus eterna

MN Manuscripts

x

x x

x

x

x

x

x

x

x

x

x

x x x

x

244

Iceland only Seqv 1: Lat. fragm. 418 Seqv 12: Lat. fragm. 952 Seqv 30: Lat. fragm. 236 Iceland only Seqv 1: Lat. fragm. 418 Seqv 47 add: Lat. fragm. 1075 Seqv 1: Lat. fragm. 418 Seqv 54 add: Arendal, AA 72 Seqv 1: Lat. fragm. 418 Seqv 50 add (Gr 45): Sand. kat. LR pk 692 Seqv 24 (Gr 24c): Lat. fragm. 843 Seqv 48 add: Lat. fragm. Box 44, C Iceland only Seqv 48 add: Lat. fragm. Box 44, C Iceland only Iceland only Seqv 12: Lat. fragm. 952 (Gjerløw id.) Seqv 12: Lat. fragm. 952 Seqv 22a: Lat. fragm. 1007 Seqv 41 add (Gr 20): Lat. fragm. 797 Seqv 1: Lat. fragm. 418 Seqv 5: Lat. fragm. 627 Seqv 36 add: Lat. fragm. 573 Seqv 12: Lat. fragm. 952 Seqv 13: Lat. fragm. 986 Seqv 27: Lat. fragm. 870 Seqv 29: Lat. fragm. 251 Seqv 32 add: Lat. fragm. 261 Seqv 37 add: Lat. fragm. 664

Indices Title Si vis vere gloriari

AH

Egg. Gj. ON

8:15

x

Sollempnitas sancti Pauli Sospitati dedit egros Spe mercedis et corone Sponsa Christo dilecta Stans a longe Stella maris o Maria exp. Stella maris o Maria pred. Stirpe Anna regia Stirpe Maria regia190 Stola iocunditatis

53:205 40:298 55:9 40:176 53:158 54:283 37:93 39:81 40:147 53:95 54:61

x

Summi regis archangele

53:192

x

Summi triumphum regis

53:67

x

Superne matris gaudia Surgit Christus cum troph. Surgit mundi vergente Tibi cordis in altari Unus amor et una conc.

55:37 54:230

x x

37:152 54:279 8:281 54:79 39:184

x x x x

x x x x x x x

Ut auctoris testatur (Ut leonis...) Uterus virgineus Veneremur crucis lignum Veni sancte spiritus

54:248 54:129

x x

54:153

x

Veni spiritus eternorum Verbum bonum et suave

53:71 54:218

x x

Victime paschali laudes

54:7

x

Virgini Marie laudes (inc. only)

54: 18 or 54:21 54:21

x

54:18

x

53:246

x

40:246

x

Virgini Marie laudes conc. Virgine Marie laudes inton. Virginis venerande

Virgo mater gratuletur

x

x

MN Manuscripts

x

x x x

x x x

x

x x

x

Seqv 60 add: Cph KB, MS NKS 133 f, 4° See Regis et pontificis, and Eggen I, p. 124. Iceland and MN only Seqv 43 add (Ant 34): Lat. fragm. 803 Seqv 8: Lat. fragm. 506 Iceland only Iceland only Seqv 14: Lat. fragm. 978 Seqv 1: Lat. fragm. 418 Seqv 15: Lat. fragm. 419 Seqv 17 (Gr 10): Bergen, SA, Rosendal Seqv 56a add: Bergen, UB, MS 1550,7 Seqv 36 add: Lat. fragm. 573 Seqv 38 add (Br-Mi 3): Lat. fragm. Sk 288 Seqv 17 (Gr 10): Bergen, SA, Rosendal Seqv 20: Lat. fragm. 1114 Seqv 30: Lat. fragm. 236 Seqv 32 add: Lat. fragm. 261 Iceland and MN only Seqv 12: Lat. fragm. 952 Seqv 26: Lat. fragm. pk. no 52 Seqv 14: Lat. fragm. 978 Iceland only

x

Seqv 4: Lat. fragm. 980 x

x x

x

x

x

x

x x

x

x

190

Iceland only, see Eggen 1968 I, p. LIV Seqv 24 (Gr 24c): Lat. fragm. 843 Seqv 61a add.: DRA, Bahus 1622-23 Seqv 1: Lat. fragm. 418 Seqv 13: Lat. fragm. 986 Seqv 14: Lat. fragm. 978 Iceland only Seqv 19: Lat. fragm. 634 Seqv 20: Lat. fragm. 1120 Seqv 30: Lat. fragm. 236 Seqv 45a add (Ant 18): Lat. fragm. 984 Seqv 46 add: Lat. fragm. 1029 Seqv 18 (Mi 106): Lat. fragm. 882 Possibly last word, Seqv 17: Bergen, SA, Rosendal Iceland and MN only Seqv 1: Lat. fragm. 418 Seqv 12: Lat. fragm. 952 Seqv 13: Lat. fragm. 986 Seqv 14: Lat. fragm. 978 Seqv 61a add: DRA, Bahus a 1622-23 Iceland only

This sequence is transmitted in one fragment only, which is most likely brought to Norway after the Reformation.

245

Title

AH

Egg. Gj. ON

MN Manuscripts

Voce iubilantes magna

10:37

x

x

x

Zyma vetus expurgetur

54:149

x

x

x

246

Seqv 1: Lat. fragm. 418 Seqv 13: Lat. fragm. 986 Iceland and MN only

Indices

Contents of Norwegian manuscripts with sequences Gjerløw’s catalogue numbers, signatures

Sequence titles (in alphabetical order)

Seqv 1 (Lat. fragm. 418):

Ad celebre rex celice Adest nobis dies alma Alle celeste nec non Aurea virga Benedicta sit beata Christo inclita candida Clare sanctorum senatus Congaudentes exultemus Iohannes Iesu Christo Iubilemus omnes una Laudes crucis attollamus Laudes deo devotas Laurenti David magni Lux illuxit letabunda Magnus Deus Mane prima sabbati Nato canunt omnia Qui regis sceptra Quicunque vult salvus esse Regnantem sempiterna Rex omnipotens Sacerdotum Christi Martinum Sacrosancta hodierna Salus eterna Sancti baptiste Stella maris o Maria pred. Veni sancte spiritus Virginis venerande Voce iubilantes magna

Seqv 2 (Lat. fragm. 366):

Celsa pueri Christi hodierna Eia recolamus Iubilemus omnes una Nato canunt omnia Qui regis sceptra

Seqv 3 (Lat. fragm. 284):

Adest nobis Ecce pulchra Organicis canamus

Seqv 4 (Lat. fragm. 980):

Mane prima sabbati Ut leonis testatur Eia recolamus Iohannes Iesu Christo Letabundus exultet

Seqv 4 (Lat. fragm. 1101):

Seqv 5 (Lat. fragm. 627):

Ad celebres rex celice Benedicta semper sancta Clare sanctorum senatus Congaudentes exultemus

247

Gjerløw’s catalogue numbers, signatures

Sequence titles (in alphabetical order) Festa Christi omnis Laudis crucis attollamus Letabundus exultet Leta mundus sit iocundus Petre summe Christi pastor Sancti baptiste

Seqv 6 (Lat. fragm. 911):

Christi hodierna Mundo Christus oritur Pangat nostra concio

Seqv 7 (Lat. fragm. 1088):

Celsa pueri concrepent

Seqv 8 (Lat. fragm. 506):

Celsa pueri concrepent Spe mercedis et corone Exulta celum Pangat nostra concio Petre summe Christi pastor

Seqv 8 (Lat. fragm. 507):

Seqv 9 (Lat. fragm. 533):

Fulgens preclara Gaudete vos fideles In sapientia disponens omnia

Seqv 10 (Lat. fragm. 427):

Ave Maria gratia plena Missus Gabriel de celis

Seqv 11 (Lat. fragm. 497):

Fulgens preclara

Seqv 12 (Lat. fragm. 952):

Adest nobis dies Benedicta sit beata Clare sanctorum senatus Lauda Syon Laus devota mente Rex omnipotens Sanctarum virginum Sancte Syon assunt Sancti spiritus assit Victime paschali (Eggen: Surgit Christus cum tropheo) Virginis venerande

Seqv 13 (Lat. fragm. 986):

Benedicta sit beata Laudes crucis attollamus Lux illuxit...lux illustris Mane prima sabbati Pangat nostra concio Predicasti dei care Regis et pontificis Sancti spiritus assit Veni sancte spiritus Virginis venerande Voce iubilantes magna

Seqv 14 (Lat. fragm. 978):

Stella maris o Maria exp. Tibi cordis in altari Veni sancte spiritus

248

Indices Gjerløw’s catalogue numbers, signatures

Sequence titles (in alphabetical order) Virginis venerande

Seqv 15 (Lat. fragm. 419):

Alle celeste nec non Ave mundi spes Maria Congaudentes exultemus Ecce pulchra Stella maris o Maria pred.

Seqv 16 (Lat. fragm. 471):

Christo inclita candida Congaudentes exultemus

Seqv 17 (SAB, Rosendal dok.):

Psallat ecclesia m. illibata Stirpe Anna regia Summi regis archangele

Seqv 18 (Lat. fragm. 881):

Grates honos hierarchia Martiris eximii Petre summe Christi pastor Sollemnitatem sancti Pauli Virginis [sic] Marie laudes (inc. only) Lux illuxit letabunda

(Lat. fragm. 882): (Lat. fragm. 932): Seqv 19 (Lat. fragm. 634):

Ave mundi spes Maria Missus Gabriel de celis Verbum bonum et suave

Seqv 20 (Lat. fragm. 1114):

Laudes crucis attollamus Summi regis archangele Inviolata Verbum bonum et suave

(Lat. fragm. 1120): Seqv 21 (Lat. fragm. 715):

Alle cantabile Interni festi gaudia

Seqv 22a (Lat. fragm. 1007):

Inviolata

Seqv 22b (Lat. fragm. XIX):

Sancte Syon assunt

Seqv 23 (Lat. fragm. 281):

Aurea virga

Seqv 24 (Lat. fragm. 843):

Organicis canamus Psallat ecclesia m. illibata Salve crux sancta Veneremur crucis lignum Virginis venerande

Seqv 25 (Lat. fragm. 532):

Pangat nostra concio Quare fremuerunt gentes unidentified

Seqv 26 (Lat. fragm. pk. 52 (X)):

Surgit mundi vergente

Seqv 27 (Lat. fragm. 870):

Sancti spiritus assit

Seqv 28 (Lat. fragm. 694):

Celi enarrant in terram Dilecto regi virtutum

249

Gjerløw’s catalogue numbers, signatures

Sequence titles (in alphabetical order) Gratulemur dulcis prosa Iocundare plebs fidelis Mundi etate octava Nunc nobis succurre

Seqv 29 (Lat. fragm. 251):

Alma chorus domini Sancti spiritus assit

Seqv 30 (Lat. fragm. 236):

Agni paschalis esu Laudes Christo redempti Rex omnipotens Summi triumphum regis

Seqv 31 add (Lat. fragm. 235):

Omnes sancti seraphim

Seqv 32 add (Lat. fragm. 261):

Sancti spiritus assit Summi triumphum regis

Seqv 33a add (Lat. fragm. 294):

Alle celeste nec non

Seqv 33b add (Lat. fragm. 489):

Celeste organum (inc. only) Clare sanctorum (inc. only) Ecce pulchra (inc. only) Eia recolamus (inc. only) Pro nobis ora (inc. only)

Seqv 34 add (Lat. fragm. 553):

Benedicta semper sancta Lauda Sion salvatorem

Seqv 35 add (Lat. fragm. 571):

Eia recolamus Hanc concordi famulatu Iohannes Iesu Christo

Seqv 36 add (Lat. fragm. 573):

Congaudent angelorum Petre summe Sancti baptiste Christi prec. Stola iocunditatis

Seqv 37 add (Lat. fragm. 664):

Resonet sacrata iam Sancti spiritus assit

Seqv 38 add (Lat. fragm. 668 ao):

Alle cantabile Potestate non natura Stola iocunditatis

Seqv 39a add (Lat. fragm. 750):

Celi enarrant Christi bapt. Petre summe

Seqv 39b add (Lat. fragm. 764):

Gaudia mundo (inc. only)

Seqv 40 add (Lat. fragm. 776):

Nunc luce alma

Seqv 41 add (Lat. fragm. 797):

Sancte Syon assunt

Seqv 42 add (Lat. fragm. 798):

Congaudentes exultemus

250

Indices Gjerløw’s catalogue numbers, signatures

Sequence titles (in alphabetical order)

Seqv 43 add (Lat. fragm. 803):

Sospitati dedit egr.

Seqv 44 add (Lat. fragm. 813):

Epiphaniam domini O mira domini

Seqv 45a add (Lat. fragm. 984ab):

Victime paschali

Seqv 45b add (Lat. fragm. 1028):

Eya recolamus

Seqv 46 add (Lat. fragm. 1029):

Ab hac familia Victime paschali

Seqv 47 add (Lat. fragm. 1075):

Congaudentes Deus in una Sacerdotem

Seqv 48 add (Lat. fragm. box 44, C):

Salve crux sancta arbor Salve martyr alme

Seqv 49 add (Lat. fragm. pk. 717):

Sancte Sion assunt

Seqv 50 add (Lat. fragm. pk. 692):

Salve crux arbor

Seqv 51 add (Lat. fragm. pk. 157):

Iocundare plebs fidelis

Seqv 52 add (Lat. fragm. pk. 718):

Mane prima sabbato

Seqv 53 add (Lat. fragm. pk. 142):

Celica resonant Nato canunt Natus ante secula

Seqv 54 add (Arendal, AA 72):

Congaudentes exultemus Psallat ecclesia mater illibata Sacrosancta hodierna

Seqv 55 add (Arendal, AA 4981-2):

Festa Christi omnis Dixit dominus ex basan unidentified

Seqv 56a add (Bergen, UL, MS 1550,7):

Stirpe Maria regia

Seqv 56b add (private ownership):

Six sequences191

Seqv 57 add (MCH, Løsfunn Lom A36 a):

Aquas plenas

Seqv 58 add (Trondh. GB, fragm. 29):

Benedicta semper sancta Profitentes unitatem Quicumque vult salvus esse

Seqv 59a add (Cph KB, MS Add 120 fol.):

Festa Christi omnis Laudes salvatori

191

These sequences were not included in Gjerløw’s list of sequences found after 1957, presumably because the fragments have no connection to medieval Norway.

251

Gjerløw’s catalogue numbers, signatures

Sequence titles (in alphabetical order)

Seqv 59b add (Cph KB, MS NKS 32 8°):

Hec est sancta sollemnitas

Seqv 60 add (Cph KB, MS NKS 133 f, 4°:

Ave mundi spes Maria Benedicta es celorum Sancti spiritus assit

Seqv 61a add (Cph DRA, fragm. 3031-3032):

Ad laudes salvatoris Ave preclara maris stella Missus Gabriel de celis Veneremur crucis lignum Virgines venerande

Seqv 61b add (Cph DRA, fragm. 3071-3072):

Agone triumphali Clare sanctorum senatus Interni festi gaudia Mane prima sabbati

Seqv 62 add (Stav. museum library, Scriver):

Fulgens preclara Prome casta contio

Seqv 63 add (Lat. fragm. 780):

Inviolata, intacta et casta

Seqv 64 add (Lat. fragm. 1030):

Lux illuxit (inc. only)

252

Indices

Sequences through the church year Proprium de tempore:

Ordo Nidrosiensis Ecclesiae

Missale Nidrosiense

Dominica I adventus BMV (tempore adventus) Dom. II adventus Dom. III adventus Dom. IV adventus Natale dni ad primam missam Natale dni ad secundam missam Natale dni ad tertiam missam Stephani protomartyris (26 Dec.) BMV tempore nativitatis Iohannis ev. (27 Dec.) Innocentium (28 Dec.) Thome Cantuarien. (29 Dec.) Sexto die nat. si dominica Silvestri pape (31 Dec.) In circumcisione dni (1 Jan.) Octava Stephani protom. (2 Jan.) Octava Iohannis ev. (3 Jan.) Octava Innocentium (4 Jan.) Vigilia Epiphanie (5 Jan.) Per oct. Epiphanie (6 Jan.) Feria iv et vi post Epiph. Octava Epiphanie Dom. I post oct. Epiph. Dom. II Dom. III Dominica resurrectionis Dom. resurrect. ad vesperas BMV in tempore paschali Feria ii post resurrect. Feria iii Feria iv Feria v Feria vi Sabbato post resurrect. Dominica in albis Dom. II post pascha Dom. III Dom. IV Dom. V Ascensionis dni Dom. post ascensionem Dom. pentecosten

Salus eterna Regnantem sempiterna Gaudia mundo Iubilemus omnes una Nato canunt omnia Celeste organum vel Lux fulget Celica resonant Magnus deus Virgo mater Celsa pueri Spe mercedis Lux fulget hodierna Pangat nostra concio Eia recolamus Hanc concordi famulati Iohannes Iesu Christo Laus tibi Christe cui Letabundus Epiphaniam Domino O mira domini pietas Festa Christi In sapientia disponens omnia In sapientia... In sapientia… Fulgens preclara Victime paschali laudes Zima vetus Prome casta concio Laudes Christo redempti Dic nobis quibus Sancta cunctis letitia Pangamus creatoris Laudes salvatori Laudes Christo redempti Sancta cunctis letitia Pangamus creatoris Zima vetus Rex omnipotens Summi triumphum regis Sancti spiritus assit

Salus eterna Missus Gabriel de celis

Feria ii post pent. Feria iii Feria iv Feria v Feria vi Sabbato post pentecost.

Resonet sacrata Eia musa Almiphona iam gaudia Veni spiritus eternorum Alma chorus domini Laudes deo devotas

253

Grates nunc omnis reddamus Celeste organum Letabundus exultet

Eia recolamus

Epiphaniam Domino In sapientia… Victime paschali Virgini marie laudes

Rex omnipotens die hodierna De sancto spiritu (rubr): Sancti spiritus assit. Alia: Veni sancte spiritus

Proprium de tempore:

Ordo Nidrosiensis Ecclesiae

Missale Nidrosiense

Dom. trinitatis

Benedicta sit beata

Dom. I post trinitatis De Corpore Christi (from 1311) Dom. II Dom. III Dom. IV Dom. V Dom. VI Dom. VII Dom. IX Dom. X Dom. XI Dom. XII Dom. XIII Dom. XIV Dom. XV Dom. XVI Dom. XVII Dom. XVIII Dom. XIX Dom. XX Dom. XXI Dom. XXII Dom. XXV

Benedicta semper

Benedicta semper sancta Alia: Voce iubilantes magna Alia: Iubilemus cordis voce Alia: Quicumque vult salvus esse

O alma trinitas Voce iubilantes Iubilemus cordis Benedicta sit beata Benedicta semper sancta Voce iubilantes Iubilemus cordis Benedicta sit beata Stans a longe Benedicta semper sancta O alma trinitas Voce iubilantes Iubilemus cordis Benedicta sit beata Benedicta semper sancta O alma trinitas Voce iubilantes Iubilemus cordis Benedicta sit beata Benedicta semper sancta O alma trinitas

Proprium de sanctis:

ON

Andree ap. in die (30 Nov.) Barbare v.m. (4 Dec.) Nicholai Myrensis (6 Dec.) Octava Andree (7 Dec.) Lucie v.m. (13 Dec.) Thome apostoli (21 Dec.) Fabiani et Sebastiani (20 Jan.) Agnetis v.m. (21 Jan.) Vincentii martyris (22 Jan.) Babille (24 Jan.) In conversione Pauli (25 Jan.) Polycarpi (26 Jan.) Agnete (nativitas) (28 Jan.) Brigide (1 Feb.) In purificatione BMV (2 Feb.)

Sacrosancta hodierna Virgines venerande Congaudentes Precluis ecclesia Regem celi cantico Clare sanctorum Ecce pulchra Pure mentis gaudia Precelsa seclis colitur Dixit Dominus ex basan Claris vocibus vel Concentu parili Laudes debitas -

Blasii (3 Feb.) Agathe (5 Feb.) Vedasti et Amandi (6 Feb.) Valentini (14 Feb) Iuliane (16 Feb) In cathedra Petri ap. (22 Feb.) Mathie (24/25 Feb.) Gregorii (12 March)

Hodierne lux diei

MN

254

Concentu parili

Indices Proprium de sanctis:

ON

Cuthberti (20 March) Benedicti (21 March) Annuntiatio BMV (25 March) Tiburtii et Valeriani (14 April) Georgii (23 April) Marci (25 April)

Nunc nobis Clare sanctorum vel Mundi etate Clara chorus vel Quam dilecta vel Sancte Syon Laudes crucis attollamus Virgo mater Mundi etate, add. Lux illuxit...lux est nobis Hodierna resonent Christo regi laudes Pangat nostra concio Exulta celum

Vitalis (28 April) Dedic. eccl. (Nidr.) (29-31 April) Philippi et Jacobi (1 May) Inventio st crucis (3 May) Alexandri soc. (3 May) Iohannis ad port. lat. (6 May) Iohannis Beverlacens. (7 May) Gordiani et Epimachi (10 May) Nerei a.o. (12 May) Hallvardi martyris (15 May) Dunstani Cantuariens. (19 May) Urbani papae (25 May) Augustini Anglorum (26 May) Petronille (31 May) Marcellini et Petri (2 June) Bonifatii (5 June) Medardi et Gildardi (8 June) Primi et Feliciani (9 June) Barnabe ap. (11 June) Basildis Cirini (12 June) Basilii (14 June) Viti et Modesti (15 June) Botulphi abb. (17 June) Marci et Marcelliani (18 June) Gervasii et Prothasii (19 June) Leofredi abb. (21 June) Albani (22 June) Iohannis baptiste in die (nat.) (24 June) Iohannis bapt. in crastino (25 June) Iohannis et Pauli (26 June) In commem. Ioh. bapt. (27 June) Petri et Pauli ap. (29 June) Comm. Pauli apostoli (30 June) Iohannis baptiste in oct. (1 July) Suithuni ep. cf. (2 July) Processi et Martiniani (2 July) In crastino Suithuni (3 July) In transl. Martini ep. cf. (4 July) Octava apostolorum (6 July) Sanctorum in Selio (8 July) Septem fratrum (10 July) In translatione Benedicti abb. (11 July) In divisione apostolorum (15 July)

MN

Sancti baptiste Christi Agone triumphali Sancti baptiste Christi Petre summe Solennitas sancti Pauli Exulta celum Psallat ecclesia vel Pangat nostra concio Agone triumphali Laude iocunda Sacerdotem Christi Petre summe Ecce pulchra Sancti merita benedicti vel Laudent eum Celi enarrant

255

Sancte Syon assunt

Sancti baptiste Christi

Proprium de sanctis:

ON

Suithuni ep. cf., transl. (15 July)

Psallat ecclesia vel Pangat nostra concio Hac in die magnalia Mane prima vel Laus tibi Christe Organicis vel Ut auctoris testatur Lux illuxit

Margarete v.m. (20 July) Praxedis (21 July) Marie Magdalene (22 July) Apollinaris (23 July) Iacobi apostoli (25 July) Septem dormientium (27 July) Felix a.o. (28 July) Olavi regis et martyris (29 July) Abdon et Sennen m. (30 July) Petri ad vincula (1 Aug.) Stephani pape martyris (2 Aug.) Inventionis Stephani alior. m. (3 Aug.) In transfiguratione domini (5 Aug.) Sixti a.o. (6 Aug.) Donati (7 Aug.) Cyriaci a.o. (8 Aug.) Laurentii in die (10 Aug.) Tiburtii (11 Aug.) Laurentii in crast. (11 Aug.) Ypoliti soc. m. (13 Aug.) Assumptionis BMV (15 Aug.) In crastimo assumptionis (ii) (16 Aug.) Octava Laurentii (iii) (17 Aug.) feria iv post assump. (18 Aug.) feria v (19 Aug.) Sabbato (21 Aug.) Octava assumpt. (22 Aug.) Bartholomei ap. (24 Aug.) Audoeni (24 Aug.) Hermetis (27/28 Aug.) Augustini Hipponensis (27/28 Aug.) Decollatio Iohannis (29 Aug.) Felicis et Adaucti (30 Aug.) Egidii abbatis (1 Sept.) Transl. Cuthberti ep. (4 Sept.) Berthini (5 Sept.) Adriani (7 Sept.) In natale BMV (8 Sept.) Gorgonii (9 Sept.) Audomari (9 Sept.) Prothi et Iacincti (11 Sept.) Cornelii et Cypriani (14 Sept.) In exaltatione st crucis (14 Sept.) Nicomedis (15 Sept.) Eufemie (16 Sept.) Lanberti (17 Sept.) Mauritii a.o. (22 Sept.) Firmini (25 Sept.)

MN

Agone triumphali Nunc luce alma Unus amor Ecce pulcra Fulget mundo celebris Martyris eximii Adest nobis Laurenti David Agone triumphali A rea virga Congaudent angelorum Stola iocunditatis Ave preclara Potestate non natura Post partum A rea virga Alle cantabile Interni festi gaudia Exultet in gloria Alme confessor (ant.) Alle celeste (see 16.+ 18.-22. August) Grates honos Alludat -

256

Mane prima sabbati

Lux illuxit Alia: Postquam calix hodierna

A rea virga

Nativitas marie virginis

Indices Proprium de sanctis:

ON

MN

Cosme et Damiani (27 Sept.) Michaelis archangeli (29 ept.) Hieronymi (30 Sept.) Remigii germani (1 Oct.) Leodegarii (2 Oct.) Fidis (6 Oct.) Marci pape (7 Oct.) Dionysii (9 Oct.) Gereonis (10 Oct.) Calixti pape (14 Oct.) Luce ev. (18 Oct.) Undecim milium virginum (21 Oct.) Severini ep. Colonien. (23 Oct.) Simonis et Iude (28 Oct.) Quintini (30 Oct.) Omnium sanctorum (1 Nov.)

Christo inclita candida

Omnium fidelium defunctorum (2 Nov.) Eustachii (2 Nov.) Leonardi ep. (6 Nov.) Quatuor coronatorum (8 Nov.) Theodori (9 Nov.) Menne (10/11 Nov.) In natali Martini ep. (11 Nov.) In crastino... (12 Nov.) Bricii ep. cf. (13 Nov.) Edmundi regis martyris (20 Nov.) Cecilie v.m. (22 Nov.) Clementis pape (23 Nov.) Chrysogonis (24 Nov.) Katherine v.m. (25 Nov.) In festo reliquiarum Comm. s. crucis feria sexta De apostolis Alia de apostolis De evangelistis De uno martyre De martyribus Alia sequentia De uno confessore De una virgine In commemoratione s. crucis Alia sequentia In commem. BMV Alia de beata virgine Alia sequentia De angelis

Mera promat uranica Sacerdotem Christi Martinum Sacerdotem... Pangat nostra concio Omnis fidelium ecclesia Sponsa Christo Festa dies Dilecto regi Superna matris gaudia Salve crux sancta

257

Christo inclita gaudia Alia: Superne matris gaudia

Alleluia nunc decantet Clare sanctorum senatus Laus devota mente Adest nobis dies alma Agone triumphali militum O beata beatorum martyrum Pangat nostra contio Laudes debitas deo nostro Veneremur crucis lignum Salve crux sancta arbor Ave preclara maris stella Hodierne lux diei Celum dus inclinavit Ad celebres rex celice

Bibliography

! $

% .

2 999

%

5 6 7 :

/4 < = 3 @

!!)

& '( 0 * % 1,

/ " # 3 /4 $

"

% A

3

# *+ ,

- #

8 6

;

*) # *> ?

& ?

999

5 6 + 1 B

1

*?

' A <8 # > 3 * + # 3 4 '

#

7

(' ) 44C;

)

/ ! E

#

# 1

0

, G

*! <

3 > D = *

1

+ +!!

8

. '/ ' F D < > > #* B 1 * F B < 999 C : G6 = > A > # = $ , ', . &! / 0 1 *A # + # G G < # # 3 + # $ , * , ) * D* . C 2 G < < 1 H I ? D 8 4/ 3 4 %+!!566 7 3 7? J: ## 1 8 , / 3 F ; 1 # ; C # ! , 1 * + 2!!2 5 29 :;<=;%, 2; = D* 6 3 ! # # < . 44 C/ ;C 5 ) *G A * A + C' / ! 1 0 * %1 G > < = =

258

Bibliography >

7-

.3 5

. ! 999

733

CC 0 /

999

7-

0 # 1 : ?

15 33 1

'

3

*

# B C

!

( . !

?

# !

)

1

*?

*

! ?

%, * !

: B

! "

7 5?

.3

. ?

3 )

0

%" 1

5> 33 1 : 5? )

C'

!

5

- )%@ .3 5

F #

K H # ', #

' ? K # < > / 6 5 . &! / + # G <

%A

: 1 = $ , 0 1 *A # 3 )1 . L# 5 1 $ (, )/ 4;' ! 8 - # '' 1 . '' ' A <8 # ! + * B ? / . ' ' A <8 # 999 ' B ? C # / '' 1 . ' ' A <8 # ! 3 > D 1 C . / ' 5 6 7 0 < ! < 6 ! 1 > H : 0 = ? D B ! * 3 # + 5M G < # # 999 C : 3 # 0 < 5 = $ , ', . &! / 0 1 *A # + # G < # 999 C 3 : *+ ! 5 G = $ , ', . &! / 0 1 *A # + # G < # * A D < = * > ? 6 * * B 1 *< * A ) E+ 1 E2' 1 0 , ' ?

259

6

> A

'

? A

'

! << ! D

< # D

' ' 0 ! +

' 4 , A ?

;F==

*

5 999

-

9

%9

#

;

2+

GH .

'

// $

-

* B

4 ; 4'C $ % -

1

* % '1

8

+ D E

# N 1

4 * D

I

# N

$ ? > $ / 1 0 ! # ! # B 1 *? D H # # . ' 6 ; 3 # D H # # D A 444 D , 9 , > # 8 , (8,) 4 ; ! D* , 8 ;J ;K=J5;GJJ L : ;M G G < )## ) C4 $ ' , N . OO= O O == ' O + > # ) =# A % > , E 1 > G ? ) 1 44' 8 5, B B O ;G # > + ' 5 ? 0 . ! # ! # B 1 *? H $ 0 0 : 2;K; 8 > ! ? HN + # C , (? N) 1 26!!5 6!+ ) *5 % ; ===1 . =B , F B 1 # 3 8 ? + +K @ # #> 4 P 0 / B # ' ) *@ + 8 # # . '4 ' # ' @ .+ ! + D

8

260

Bibliography 3% E6 A 999

' D

<

< # * C 4'; 4 = 9 *H +E

'C Q

1

E F

#

<

E

3*

# 999

'

; #

'

; 1 ' ' ;C/ .

Q C '

999

#

L

0

8 0

B

1 C4 *

999 = B

? , ===1

. #

1 / >

999

/ == B

5

= 1

999

#

B

/ > 5 G

1

1

// *H +E 999 /4 G 1 # F 999 / ' , 999 44 G 3 > ) # ' 3 # ? 4/4 D 3 > 0' / $ 999 ' ? ) # 5

999 3 999 3 )

3* #D 3E P , F

:

*5 +

%

,

; F

'

E

% I

# < E I <==

%

A

= S

*?

*H

F

;

#

4 G

/

F 6

<

# ; D %

! 0

6

: 0

# 2'

*

1

2

7 0

0 1

!

5 H

=== B

.

44 ; 0 H #

#

*

*H

E

:

G G 5 1 / B

# !! H ** 4/

F

: 6 #

E F

D T==K

?

B 5

5

# = 9

999

, =

<

F = *

E B

,

( R N )) ?

%E #1 Q I

'

#

#

1 :

G &( < &

2 = <

*! # ; C # (= ;K5;F

P @

!

2!

#* ' 5 # G

P8 1 ;E 2S

261

; 11 ;4

# D

;

999

G

H

## C /

#

+

> F

999

D /

F

# )1

. L#

5

G * +

:

) D

#

;G

%9

999

/

1

;!

*>

=

0 0

# - #

G 2

: 2 ;T 5 ! 2 # D 8 )

) B

,

G

H

2 1 -

3

# ; = 8 0

# +K 1;T; *F A # @ :

# #

1 < *F

$P ;. #

8 =

4

< = 5 6 * = $ 5 6 7 @ 3 * # # + 0

A

I

+

#

@ F

999 + 8

#

/ >

1

4/ >

#

1

;

U

#

+

*

0

)2

999 3 ) B 1 + # 0 # '/;' + )

>

# #

< & ( 4; ' '/ * B %" "

H # #

) !@

A

*Q

<

+

D ;;2 5;K2

GE # + 0 + + * D 1 +% F

E

= ;<
*

#0 0

% * ) 3 ) B # 1 # C 0

1 ? -1

F2 , F +%

F + ;F A 1# F @ ! << H + 'C 999 'C B * H!!0 F < #1 + G 4C *

E GE GE

# #

0 !

;T2

? >

Q O #

2 + I

F @

@

3* 8

, 1#

2V

: 2T

%

I

L * #

262

R

G +

S

Bibliography 999

1 < *) > +

# 999 ?

*#

C D

+ #

6

1 1 T= , =U* # #
< ## = + 1 1 = . *) >

+

999

#

=

E 3 %

*0

%E 5

O

0 *

# ; #

B 1 - )

, @ C + C , 0 #1

*

*

G *A 3 ;

G

? .

;MG= 3E

;KG=

=

;K== )

%

*

,

# : 2T F

%D

1#

L = D ) 1

# ;
<2Q ' D

B P )

1

5* # 2

# (> F B 1 /' F %

20 2'

CC &

F;
# 3

<

*

$ @ 2HT )

* %

5

=

1

D* 6 C 0

) ) <

. C '4;44 999 ' A 8 # M &; ( ; /

<

:F; I

'

# O

I

* %8 D

E #

*

W

F ) ;

(, F @

D

44 B

1

1

#D

G

> $ +V > # ' + N# > D* 6 +E E H L < 999 4 . J=5 =

?L <

4

# + #

G* % 2D

2 2 # 1

# 2. #

/ =

E

5

#

#

2. 2

+

# Q 2' L * =

<

5

263

# <

;

< G

#

N

# #

< G <

%

%

=

999

G

#

1

=

<

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

Bibliography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

>

265

F

> #

0

' , '1 . 999 '/ ,

< TT

*

999

) , ,

.

Q ' G

# # H #

? # C G

999 @ ! 999

# ?

<

F

5 H #

! #

' ; 4;

:

? * 4 $

>

<

0 <

F

'

A

# <

% C ) <

3

' ! # B 1

# . ;'4

#

< ?

* 5Z

!

? F2

# 1

# 2.

Q 2' #

# Q 2' /C D Z

? 6 8 3

6 * = $ # G

5 +

# /

3

3 +

4;

;

/ /4;4' ' ; @ ' @ <

!

*A #

>

?

1

5 #

#

G6

,

? ?

E

;

1 C

?

.

G #

% F % F

*5

' ,

.

%

. )

'1

0

.

C ,

F 999

*5 F *5

) '1

999

0* *

# <

# 1 .

% C

;

2. ;44

C *

# 2T%

2 5

2

, :;<;; 2 G

=

D* 6 0

)1

/ ? O

8 ! <

0V

?

? < D . !

999

< 4 , '' ,

# >

B

O

1 %

*

:T # ' ? O 2 2, B << +

! < %. # G ! ) 8

;KJ=

266

#

# % *

F

E +

;5 #P #

Bibliography ' >

999

%

, # !

*

2V

]! = $

1

-

, # 4 ,

1 ? . 1

3

3 ',

*A #

@

H 4 44

44' , 4 * !

K

/'

2

$

G

F2 *> 4' *

G G *

# ! 4/

)

B 1

F2 C 4+ C B # B 1

B

B G 0

4

, B # ! 0 9 5 , . 1 ) 999 C / . L#

)1

; '4;/ ,

)

! *D

*@

.

#

E

1

*? #

5

'

. #[ .

A 2 /

# 4 , % 1 B

*? 3

$ ?

50

! 2

# 267

$

;;==5;G==

D

1

F 1

/

4/; 4 # C4 ; 4C D Q D O > # * G 5

#

3 $

:

'

2S >

<

)

<

+ !

<

L 0

0

(,

5

' :TH A

>

0 # 1 > % ' K # )

G

G1

1

# ;H &! /

8 . # A] _ # 3 4 0 3 , 9 E## ;KHJ ' 1 . , O * 3E ) A <8 #

>

< /

. # G

+

/ , B

A

E E

#1

C F

G 0

1

<

L 5

999 999

' F;<- 0 : 2; F

)#

#

%

#% 1

=

@

) B

@ @

8

C D H # 6 # ` 4 4 ; # 0 " % /

8

86 #

#

V

"

# > &5 6 (?

#*

C' & ( .

2, @ U

-

; 4

:1

P @

444 $ 1

G ! )*

$ @

* %,

. B <<

B

B< Websites referred to in the text: Cantus: A Database for Latin Ecclesiastical Chant http://publish.uwo.ca/~cantus/ Codices Electronici Sangallenses (CESG): http://www.cesg.unifr.ch/en/index.htm Diplomatarium Norvegicum (DN): http://www.dokpro.uio.no/dipl_norv/diplom_felt.html. Fragment, University of Bergen: http://gandalf.aksis.uib.no Ismus (including manuscript fragments): http://ismus.musik.is MPO project, Riksarkivet i Stockholm: http://www.ra.se/ra/MPO-exempel.html Patrologia Latina. The full text database (PL): http://pld.chadwyck.co.uk/ Vocabulaire Codicologique: http://vocabulaire.irht.cnrs.fr/vocab.htm St. Laurentius, University of Lund: http://laurentius.lub.lu.se.

268

1

4